Re: [DISCUSS] KIP-110: Add Codec for ZStandard Compression (Updated)

[Dongjin Lee <do...@apache.org>]

I updated the KIP page
<https://cwiki.apache.org/confluence/display/KAFKA/KIP-110%3A+Add+Codec+for+ZStandard+Compression>
following the discussion here. Please take a look when you are free.
If you have any opinion, don't hesitate to give me a message.

Best,
Dongjin

On Fri, Aug 31, 2018 at 11:35 PM Dongjin Lee <do...@apache.org> wrote:

> I just updated the draft implementation[^1], rebasing against the latest
> trunk and implementing error routine (i.e., Error code 74 for
> UnsupportedCompressionTypeException.) Since we decided to disallow all
> fetch request below version 2.1.0 for the topics specifying ZStandard, I
> added an error logic only.
>
> Please have a look when you are free.
>
> Thanks,
> Dongjin
>
> [^1]: Please check the last commit here:
> https://github.com/apache/kafka/pull/2267
>
> On Thu, Aug 23, 2018, 8:55 AM Dongjin Lee <do...@apache.org> wrote:
>
>> Jason,
>>
>> Great. +1 for UNSUPPORTED_COMPRESSION_TYPE.
>>
>> Best,
>> Dongjin
>>
>> On Thu, Aug 23, 2018 at 8:19 AM Jason Gustafson <ja...@confluent.io>
>> wrote:
>>
>>> Hey Dongjin,
>>>
>>> Yeah that's right. For what it's worth, librdkafka also appears to handle
>>> unexpected error codes. I expect that most client implementations would
>>> either pass through the raw type or convert to an enum using something
>>> like
>>> what the java client does. Since we're expecting the client to fail
>>> anyway,
>>> I'm probably in favor of using the UNSUPPORTED_COMPRESSION_TYPE error
>>> code.
>>>
>>> -Jason
>>>
>>> On Wed, Aug 22, 2018 at 1:46 AM, Dongjin Lee <do...@apache.org> wrote:
>>>
>>> > Jason and Ismael,
>>> >
>>> > It seems like the only thing we need to regard if we define a new error
>>> > code (i.e., UNSUPPORTED_COMPRESSION_TYPE) would be the implementation
>>> of
>>> > the other language clients, right? At least, this strategy causes any
>>> > problem for Java client. Do I understand correctly?
>>> >
>>> > Thanks,
>>> > Dongjin
>>> >
>>> > On Wed, Aug 22, 2018 at 5:43 PM Dongjin Lee <do...@apache.org>
>>> wrote:
>>> >
>>> > > Jason,
>>> > >
>>> > > > I think we would only use this error code when we /know/ that zstd
>>> was
>>> > > in use and the client doesn't support it? This is true if either 1)
>>> the
>>> > > message needs down-conversion and we encounter a zstd compressed
>>> message,
>>> > > or 2) if the topic is explicitly configured to use zstd.
>>> > >
>>> > > Yes, it is right. And you know, the case 1 includes 1.a) old clients'
>>> > > request v0, v1 records or 1.b) implicit zstd, the compression type of
>>> > > "producer" with Zstd compressed data.
>>> > >
>>> > > > However, if the compression type is set to "producer," then the
>>> fetched
>>> > > data may or may not be compressed with zstd. In this case, we return
>>> the
>>> > > data to the client and expect it to fail parsing. Is that correct?
>>> > >
>>> > > Exactly.
>>> > >
>>> > > Following your message, I reviewed the implementation of
>>> > > `KafkaApis#handleFetchRequest,` which handles the fetch request. And
>>> > found
>>> > > that the information we can use is like the following:
>>> > >
>>> > > 1. Client's fetch request version. (`versionId` variable)
>>> > > 2. Log's compression type. (`logConfig` variable)
>>> > >
>>> > > We can't detect the actual compression type of the data, unless we
>>> > inspect
>>> > > the `RecordBatch` included in the `Records` instance (i.e.,
>>> > > `unconvertedRecords` variable.) Since it requires some performance
>>> issue,
>>> > > it is not our option - in short, we can't be sure if given chunks of
>>> data
>>> > > are compressed with zstd or not.
>>> > >
>>> > > So, conclusion: we can return an error in the case of 1.a and 2
>>> easily,
>>> > > with the information above. In the case 1.b (implicit zstd), we can
>>> just
>>> > > return the data by do nothing special and expect it to fail parsing.
>>> > >
>>> > > Thanks,
>>> > > Dongjin
>>> > >
>>> > > On Wed, Aug 22, 2018 at 12:02 PM Ismael Juma <is...@gmail.com>
>>> wrote:
>>> > >
>>> > >> Jason, that's an interesting point regarding the Java client. Do we
>>> know
>>> > >> what clients in other languages do in these cases?
>>> > >>
>>> > >> Ismael
>>> > >>
>>> > >> On Tue, 21 Aug 2018, 17:30 Jason Gustafson, <ja...@confluent.io>
>>> wrote:
>>> > >>
>>> > >> > Hi Dongjin,
>>> > >> >
>>> > >> > One of the complications is that old versions of the API will not
>>> > >> expect a
>>> > >> > new error code. However, since we expect this to be a fatal error
>>> > anyway
>>> > >> > for old clients, it may still be more useful to return the correct
>>> > error
>>> > >> > code. For example, the Kafka clients use the following code to
>>> convert
>>> > >> the
>>> > >> > error code:
>>> > >> >
>>> > >> >     public static Errors forCode(short code) {
>>> > >> >         Errors error = codeToError.get(code);
>>> > >> >         if (error != null) {
>>> > >> >             return error;
>>> > >> >         } else {
>>> > >> >             log.warn("Unexpected error code: {}.", code);
>>> > >> >             return UNKNOWN_SERVER_ERROR;
>>> > >> >         }
>>> > >> >     }
>>> > >> >
>>> > >> > If we return an unsupported error code, it will be converted to an
>>> > >> UNKNOWN
>>> > >> > error, but at least we will get the message in the log with the
>>> > correct
>>> > >> > code. That seems preferable to returning a misleading error code.
>>> So I
>>> > >> > wonder if we can use the new UNSUPPORTED_COMPRESSION_TYPE error
>>> even
>>> > for
>>> > >> > older versions.
>>> > >> >
>>> > >> > Also, one question just to check my understanding. I think we
>>> would
>>> > only
>>> > >> > use this error code when we /know/ that zstd was in use and the
>>> client
>>> > >> > doesn't support it? This is true if either 1) the message needs
>>> > >> > down-conversion and we encounter a zstd compressed message, or 2)
>>> if
>>> > the
>>> > >> > topic is explicitly configured to use zstd. However, if the
>>> > compression
>>> > >> > type is set to "producer," then the fetched data may or may not be
>>> > >> > compressed with zstd. In this case, we return the data to the
>>> client
>>> > and
>>> > >> > expect it to fail parsing. Is that correct?
>>> > >> >
>>> > >> > Thanks,
>>> > >> > Jason
>>> > >> >
>>> > >> >
>>> > >> >
>>> > >> > On Tue, Aug 21, 2018 at 9:08 AM, Dongjin Lee <do...@apache.org>
>>> > >> wrote:
>>> > >> >
>>> > >> > > Ismael, Jason and all,
>>> > >> > >
>>> > >> > > I rewrote the backward compatibility strategy & its alternatives
>>> > like
>>> > >> > > following, based on Ismael & Jason's comments. Since it is not
>>> > >> updated to
>>> > >> > > the wiki yet, don't hesitate to give me a message if you have
>>> any
>>> > >> opinion
>>> > >> > > on it.
>>> > >> > >
>>> > >> > > ```
>>> > >> > > *Backward Compatibility*
>>> > >> > >
>>> > >> > > We need to establish some backward-compatibility strategy for
>>> the
>>> > >> case an
>>> > >> > > old client subscribes a topic using ZStandard implicitly (i.e.,
>>> > >> > > 'compression.type' configuration of given topic is 'producer'
>>> and
>>> > the
>>> > >> > > producer compressed the records with ZStandard). We have the
>>> > following
>>> > >> > > options for this situation:
>>> > >> > >
>>> > >> > > *A. Support ZStandard to the old clients which can understand
>>> v0, v1
>>> > >> > > messages only.*
>>> > >> > >
>>> > >> > > This strategy necessarily requires the down-conversion of v2
>>> message
>>> > >> > > compressed with Zstandard into v0 or v1 messages, which means a
>>> > >> > > considerable performance degradation. So we rejected this
>>> strategy.
>>> > >> > >
>>> > >> > > *B. Bump the API version and support only v2-available clients*
>>> > >> > >
>>> > >> > > With this approach, we can message the old clients that they
>>> are old
>>> > >> and
>>> > >> > > should be upgraded. However, there are still several options
>>> for the
>>> > >> > Error
>>> > >> > > code.
>>> > >> > >
>>> > >> > > *B.1. INVALID_REQUEST (42)*
>>> > >> > >
>>> > >> > > This option gives the client so little information; the user
>>> can be
>>> > >> > > confused about why the client worked correctly in the past
>>> suddenly
>>> > >> > > encounters a problem. So we rejected this strategy.
>>> > >> > >
>>> > >> > > *B.2. CORRUPT_MESSAGE (2)*
>>> > >> > >
>>> > >> > > This option gives inaccurate information; the user can be
>>> surprised
>>> > >> and
>>> > >> > > misunderstand that the log files are broken in some way. So we
>>> > >> rejected
>>> > >> > > this strategy.
>>> > >> > >
>>> > >> > > *B.3 UNSUPPORTED_FOR_MESSAGE_FORMAT (43)*
>>> > >> > >
>>> > >> > > The advantage of this approach is we don't need to define a new
>>> > error
>>> > >> > code;
>>> > >> > > we can reuse it and that's all.
>>> > >> > >
>>> > >> > > The disadvantage of this approach is that it is also a little
>>> bit
>>> > >> vague;
>>> > >> > > This error code is defined as a work for KIP-98[^1] and now
>>> returned
>>> > >> in
>>> > >> > the
>>> > >> > > transaction error.
>>> > >> > >
>>> > >> > > *B.4. UNSUPPORTED_COMPRESSION_TYPE (new)*
>>> > >> > >
>>> > >> > > The advantage of this approach is that it is clear and provides
>>> an
>>> > >> exact
>>> > >> > > description. The disadvantage is we need to add a new error
>>> code.
>>> > >> > > ```
>>> > >> > >
>>> > >> > > *It seems like what we need to choose is now so clear:
>>> > >> > > UNSUPPORTED_FOR_MESSAGE_FORMAT (B.3) or
>>> UNSUPPORTED_COMPRESSION_TYPE
>>> > >> > > (B.4).*
>>> > >> > > The first one doesn't need a new error message but the latter is
>>> > more
>>> > >> > > explicit. Which one do you prefer? Since all of you have much
>>> more
>>> > >> > > experience and knowledge than me, I will follow your decision.
>>> The
>>> > >> wiki
>>> > >> > > page will be updated following the decision also.
>>> > >> > >
>>> > >> > > Best,
>>> > >> > > Dongjin
>>> > >> > >
>>> > >> > > [^1]: https://issues.apache.org/jira/browse/KAFKA-4990
>>> > >> > >
>>> > >> > > On Sun, Aug 19, 2018 at 4:58 AM Ismael Juma <is...@gmail.com>
>>> > >> wrote:
>>> > >> > >
>>> > >> > > > Sounds reasonable to me.
>>> > >> > > >
>>> > >> > > > Ismael
>>> > >> > > >
>>> > >> > > > On Sat, 18 Aug 2018, 12:20 Jason Gustafson, <
>>> jason@confluent.io>
>>> > >> > wrote:
>>> > >> > > >
>>> > >> > > > > Hey Ismael,
>>> > >> > > > >
>>> > >> > > > > Your summary looks good to me. I think it might also be a
>>> good
>>> > >> idea
>>> > >> > to
>>> > >> > > > add
>>> > >> > > > > a new UNSUPPORTED_COMPRESSION_TYPE error code to go along
>>> with
>>> > the
>>> > >> > > > version
>>> > >> > > > > bumps. We won't be able to use it for old api versions
>>> since the
>>> > >> > > clients
>>> > >> > > > > will not understand it, but we can use it going forward so
>>> that
>>> > >> we're
>>> > >> > > not
>>> > >> > > > > stuck in a similar situation with a new message format and
>>> a new
>>> > >> > codec
>>> > >> > > to
>>> > >> > > > > support. Another option is to use UNSUPPORTED_FOR_MESSAGE_
>>> > FORMAT,
>>> > >> but
>>> > >> > > it
>>> > >> > > > is
>>> > >> > > > > not as explicit.
>>> > >> > > > >
>>> > >> > > > > -Jason
>>> > >> > > > >
>>> > >> > > > > On Fri, Aug 17, 2018 at 5:19 PM, Ismael Juma <
>>> ismael@juma.me.uk
>>> > >
>>> > >> > > wrote:
>>> > >> > > > >
>>> > >> > > > > > Hi Dongjin and Jason,
>>> > >> > > > > >
>>> > >> > > > > > I would agree. My summary:
>>> > >> > > > > >
>>> > >> > > > > > 1. Support zstd with message format 2 only.
>>> > >> > > > > > 2. Bump produce and fetch request versions.
>>> > >> > > > > > 3. Provide broker errors whenever possible based on the
>>> > request
>>> > >> > > version
>>> > >> > > > > and
>>> > >> > > > > > rely on clients for the cases where the broker can't
>>> validate
>>> > >> > > > efficiently
>>> > >> > > > > > (example message format 2 consumer that supports the
>>> latest
>>> > >> fetch
>>> > >> > > > version
>>> > >> > > > > > but doesn't support zstd).
>>> > >> > > > > >
>>> > >> > > > > > If there's general agreement on this, I suggest we update
>>> the
>>> > >> KIP
>>> > >> > to
>>> > >> > > > > state
>>> > >> > > > > > the proposal and to move the rejected options to its own
>>> > >> section.
>>> > >> > And
>>> > >> > > > > then
>>> > >> > > > > > start a vote!
>>> > >> > > > > >
>>> > >> > > > > > Ismael
>>> > >> > > > > >
>>> > >> > > > > > On Fri, Aug 17, 2018 at 4:00 PM Jason Gustafson <
>>> > >> > jason@confluent.io>
>>> > >> > > > > > wrote:
>>> > >> > > > > >
>>> > >> > > > > > > Hi Dongjin,
>>> > >> > > > > > >
>>> > >> > > > > > > Yes, that's a good summary. For clients which support
>>> v2,
>>> > the
>>> > >> > > client
>>> > >> > > > > can
>>> > >> > > > > > > parse the message format and hopefully raise a useful
>>> error
>>> > >> > message
>>> > >> > > > > > > indicating the unsupported compression type. For older
>>> > >> clients,
>>> > >> > our
>>> > >> > > > > > options
>>> > >> > > > > > > are probably (1) to down-convert to the old format
>>> using no
>>> > >> > > > compression
>>> > >> > > > > > > type, or (2) to return an error code. I'm leaning
>>> toward the
>>> > >> > latter
>>> > >> > > > as
>>> > >> > > > > > the
>>> > >> > > > > > > simpler solution, but the challenge is finding a good
>>> error
>>> > >> code.
>>> > >> > > Two
>>> > >> > > > > > > possibilities might be INVALID_REQUEST or
>>> CORRUPT_MESSAGE.
>>> > The
>>> > >> > > > downside
>>> > >> > > > > > is
>>> > >> > > > > > > that old clients probably won't get a helpful message.
>>> > >> However,
>>> > >> > at
>>> > >> > > > > least
>>> > >> > > > > > > the behavior will be consistent in the sense that all
>>> > clients
>>> > >> > will
>>> > >> > > > fail
>>> > >> > > > > > if
>>> > >> > > > > > > they do not support zstandard.
>>> > >> > > > > > >
>>> > >> > > > > > > What do you think?
>>> > >> > > > > > >
>>> > >> > > > > > > Thanks,
>>> > >> > > > > > > Jason
>>> > >> > > > > > >
>>> > >> > > > > > > On Fri, Aug 17, 2018 at 8:08 AM, Dongjin Lee <
>>> > >> dongjin@apache.org
>>> > >> > >
>>> > >> > > > > wrote:
>>> > >> > > > > > >
>>> > >> > > > > > > > Thanks Jason, I reviewed the down-converting logic
>>> > following
>>> > >> > your
>>> > >> > > > > > > > explanation.[^1] You mean the following routines,
>>> right?
>>> > >> > > > > > > >
>>> > >> > > > > > > > -
>>> > >> > > > > > > > https://github.com/apache/kafka/blob/trunk/core/src/
>>> > >> > > > > > > > main/scala/kafka/server/KafkaApis.scala#L534
>>> > >> > > > > > > > -
>>> > >> > > > > > > >
>>> https://github.com/apache/kafka/blob/trunk/clients/src/
>>> > >> > > > > > > > main/java/org/apache/kafka/common/record/
>>> > >> > > LazyDownConversionRecords.
>>> > >> > > > > > > > java#L165
>>> > >> > > > > > > > -
>>> > >> > > > > > > >
>>> https://github.com/apache/kafka/blob/trunk/clients/src/
>>> > >> > > > > > > >
>>> > >> main/java/org/apache/kafka/common/record/RecordsUtil.java#L40
>>> > >> > > > > > > >
>>> > >> > > > > > > > It seems like your stance is like following:
>>> > >> > > > > > > >
>>> > >> > > > > > > > 1. In principle, Kafka does not change the compression
>>> > codec
>>> > >> > when
>>> > >> > > > > > > > down-converting, since it requires inspecting the
>>> fetched
>>> > >> data,
>>> > >> > > > which
>>> > >> > > > > > is
>>> > >> > > > > > > > expensive.
>>> > >> > > > > > > > 2. However, there are some cases the fetched data is
>>> > >> inspected
>>> > >> > > > > anyway.
>>> > >> > > > > > In
>>> > >> > > > > > > > this case, we can provide compression conversion from
>>> > >> Zstandard
>>> > >> > > to
>>> > >> > > > > > > > classical ones[^2].
>>> > >> > > > > > > >
>>> > >> > > > > > > > And from what I understand, the cases where the client
>>> > >> without
>>> > >> > > > > > ZStandard
>>> > >> > > > > > > > support receives ZStandard compressed records can be
>>> > >> organized
>>> > >> > > into
>>> > >> > > > > two
>>> > >> > > > > > > > cases:
>>> > >> > > > > > > >
>>> > >> > > > > > > > a. The 'compression.type' configuration of given
>>> topic is
>>> > >> > > > 'producer'
>>> > >> > > > > > and
>>> > >> > > > > > > > the producer compressed the records with ZStandard.
>>> (that
>>> > >> is,
>>> > >> > > using
>>> > >> > > > > > > > ZStandard implicitly.)
>>> > >> > > > > > > > b.  The 'compression.type' configuration of given
>>> topic is
>>> > >> > > 'zstd';
>>> > >> > > > > that
>>> > >> > > > > > > is,
>>> > >> > > > > > > > using ZStandard explicitly.
>>> > >> > > > > > > >
>>> > >> > > > > > > > As you stated, we don't have to handle the case b
>>> > specially.
>>> > >> > So,
>>> > >> > > It
>>> > >> > > > > > seems
>>> > >> > > > > > > > like we can narrow the focus of the problem by joining
>>> > case
>>> > >> 1
>>> > >> > and
>>> > >> > > > > case
>>> > >> > > > > > b
>>> > >> > > > > > > > like the following:
>>> > >> > > > > > > >
>>> > >> > > > > > > > > Given the topic with 'producer' as its
>>> > 'compression.type'
>>> > >> > > > > > > configuration,
>>> > >> > > > > > > > ZStandard compressed records and old client without
>>> > >> ZStandard,
>>> > >> > is
>>> > >> > > > > there
>>> > >> > > > > > > any
>>> > >> > > > > > > > case we need to inspect the records and can change the
>>> > >> > > compression
>>> > >> > > > > > type?
>>> > >> > > > > > > If
>>> > >> > > > > > > > so, can we provide compression type converting?
>>> > >> > > > > > > >
>>> > >> > > > > > > > Do I understand correctly?
>>> > >> > > > > > > >
>>> > >> > > > > > > > Best,
>>> > >> > > > > > > > Dongjin
>>> > >> > > > > > > >
>>> > >> > > > > > > > [^1]: I'm sorry, I found that I was a little bit
>>> > >> > misunderstanding
>>> > >> > > > how
>>> > >> > > > > > API
>>> > >> > > > > > > > version works, after reviewing the downconvert logic
>>> & the
>>> > >> > > protocol
>>> > >> > > > > > > > documentation <https://kafka.apache.org/protocol>.
>>> > >> > > > > > > > [^2]: None, Gzip, Snappy, Lz4.
>>> > >> > > > > > > >
>>> > >> > > > > > > > On Tue, Aug 14, 2018 at 2:16 AM Jason Gustafson <
>>> > >> > > > jason@confluent.io>
>>> > >> > > > > > > > wrote:
>>> > >> > > > > > > >
>>> > >> > > > > > > > > >
>>> > >> > > > > > > > > > But in my opinion, since the client will fail
>>> with the
>>> > >> API
>>> > >> > > > > version,
>>> > >> > > > > > > so
>>> > >> > > > > > > > we
>>> > >> > > > > > > > > > don't need to down-convert the messages anyway.
>>> Isn't
>>> > >> it?
>>> > >> > > So, I
>>> > >> > > > > > think
>>> > >> > > > > > > > we
>>> > >> > > > > > > > > > don't care about this case. (I'm sorry, I am not
>>> > >> familiar
>>> > >> > > with
>>> > >> > > > > > > > > down-convert
>>> > >> > > > > > > > > > logic.)
>>> > >> > > > > > > > >
>>> > >> > > > > > > > >
>>> > >> > > > > > > > > Currently the broker down-converts automatically
>>> when it
>>> > >> > > receives
>>> > >> > > > > an
>>> > >> > > > > > > old
>>> > >> > > > > > > > > version of the fetch request (a version which is
>>> known
>>> > to
>>> > >> > > predate
>>> > >> > > > > the
>>> > >> > > > > > > > > message format in use). Typically when
>>> down-converting
>>> > the
>>> > >> > > > message
>>> > >> > > > > > > > format,
>>> > >> > > > > > > > > we use the same compression type, but there is not
>>> much
>>> > >> point
>>> > >> > > in
>>> > >> > > > > > doing
>>> > >> > > > > > > so
>>> > >> > > > > > > > > when we know the client doesn't support it. So if
>>> > >> zstandard
>>> > >> > is
>>> > >> > > in
>>> > >> > > > > > use,
>>> > >> > > > > > > > and
>>> > >> > > > > > > > > we have to down-convert anyway, then we can choose
>>> to
>>> > use
>>> > >> a
>>> > >> > > > > different
>>> > >> > > > > > > > > compression type or no compression type.
>>> > >> > > > > > > > >
>>> > >> > > > > > > > > From my perspective, there is no significant
>>> downside to
>>> > >> > > bumping
>>> > >> > > > > the
>>> > >> > > > > > > > > protocol version and it has several potential
>>> benefits.
>>> > >> > Version
>>> > >> > > > > bumps
>>> > >> > > > > > > are
>>> > >> > > > > > > > > cheap. The main question mark in my mind is about
>>> > >> > > > down-conversion.
>>> > >> > > > > > > > Figuring
>>> > >> > > > > > > > > out whether down-conversion is needed is hard
>>> generally
>>> > >> > without
>>> > >> > > > > > > > inspecting
>>> > >> > > > > > > > > the fetched data, which is expensive. I think we
>>> agree
>>> > in
>>> > >> > > > principle
>>> > >> > > > > > > that
>>> > >> > > > > > > > we
>>> > >> > > > > > > > > do not want to have to pay this cost generally and
>>> > prefer
>>> > >> the
>>> > >> > > > > clients
>>> > >> > > > > > > to
>>> > >> > > > > > > > > fail when they see an unhandled compression type.
>>> The
>>> > >> point I
>>> > >> > > was
>>> > >> > > > > > > making
>>> > >> > > > > > > > is
>>> > >> > > > > > > > > that there are some cases where we are either
>>> inspecting
>>> > >> the
>>> > >> > > data
>>> > >> > > > > > > anyway
>>> > >> > > > > > > > > (because we have to down-convert the message
>>> format), or
>>> > >> we
>>> > >> > > have
>>> > >> > > > an
>>> > >> > > > > > > easy
>>> > >> > > > > > > > > way to tell whether zstandard is in use (the topic
>>> has
>>> > it
>>> > >> > > > > configured
>>> > >> > > > > > > > > explicitly). In the latter case, we don't have to
>>> handle
>>> > >> it
>>> > >> > > > > > specially.
>>> > >> > > > > > > > But
>>> > >> > > > > > > > > we do have to decide how we will handle
>>> down-conversion
>>> > to
>>> > >> > > older
>>> > >> > > > > > > formats.
>>> > >> > > > > > > > >
>>> > >> > > > > > > > > -Jason
>>> > >> > > > > > > > >
>>> > >> > > > > > > > > On Sun, Aug 12, 2018 at 5:15 PM, Dongjin Lee <
>>> > >> > > dongjin@apache.org
>>> > >> > > > >
>>> > >> > > > > > > wrote:
>>> > >> > > > > > > > >
>>> > >> > > > > > > > > > Colin and Jason,
>>> > >> > > > > > > > > >
>>> > >> > > > > > > > > > Thanks for your opinions. In summarizing, the
>>> Pros and
>>> > >> Cons
>>> > >> > > of
>>> > >> > > > > > > bumping
>>> > >> > > > > > > > > > fetch API version are:
>>> > >> > > > > > > > > >
>>> > >> > > > > > > > > > Cons:
>>> > >> > > > > > > > > >
>>> > >> > > > > > > > > > - The Broker can't know whether a given message
>>> batch
>>> > is
>>> > >> > > > > compressed
>>> > >> > > > > > > > with
>>> > >> > > > > > > > > > zstd or not.
>>> > >> > > > > > > > > > - Need some additional logic for the topic
>>> explicitly
>>> > >> > > > configured
>>> > >> > > > > to
>>> > >> > > > > > > use
>>> > >> > > > > > > > > > zstd.
>>> > >> > > > > > > > > >
>>> > >> > > > > > > > > > Pros:
>>> > >> > > > > > > > > >
>>> > >> > > > > > > > > > - The broker doesn't need to conduct expensive
>>> > >> > > down-conversion.
>>> > >> > > > > > > > > > - Can message the users to update their client.
>>> > >> > > > > > > > > >
>>> > >> > > > > > > > > > So, opinions for the backward-compatibility
>>> policy by
>>> > >> far:
>>> > >> > > > > > > > > >
>>> > >> > > > > > > > > > - A: bump the API version - +2 (Colin, Jason)
>>> > >> > > > > > > > > > - B: leave unchanged - +1 (Viktor)
>>> > >> > > > > > > > > >
>>> > >> > > > > > > > > > Here are my additional comments:
>>> > >> > > > > > > > > >
>>> > >> > > > > > > > > > @Colin
>>> > >> > > > > > > > > >
>>> > >> > > > > > > > > > I greatly appreciate your response. In the case
>>> of the
>>> > >> > > > dictionary
>>> > >> > > > > > > > > support,
>>> > >> > > > > > > > > > of course, this issue should be addressed later
>>> so we
>>> > >> don't
>>> > >> > > > need
>>> > >> > > > > it
>>> > >> > > > > > > in
>>> > >> > > > > > > > > the
>>> > >> > > > > > > > > > first version. You are right - it is not late to
>>> try
>>> > it
>>> > >> > after
>>> > >> > > > > some
>>> > >> > > > > > > > > > benchmarks. What I mean is, we should keep in
>>> mind on
>>> > >> that
>>> > >> > > > > > potential
>>> > >> > > > > > > > > > feature.
>>> > >> > > > > > > > > >
>>> > >> > > > > > > > > > @Jason
>>> > >> > > > > > > > > >
>>> > >> > > > > > > > > > You wrote,
>>> > >> > > > > > > > > >
>>> > >> > > > > > > > > > > Similarly, if we have to down-convert anyway
>>> because
>>> > >> the
>>> > >> > > > client
>>> > >> > > > > > > does
>>> > >> > > > > > > > > not
>>> > >> > > > > > > > > > understand the message format, then we could also
>>> use
>>> > a
>>> > >> > > > different
>>> > >> > > > > > > > > > compression type.
>>> > >> > > > > > > > > >
>>> > >> > > > > > > > > > But in my opinion, since the client will fail
>>> with the
>>> > >> API
>>> > >> > > > > version,
>>> > >> > > > > > > so
>>> > >> > > > > > > > we
>>> > >> > > > > > > > > > don't need to down-convert the messages anyway.
>>> Isn't
>>> > >> it?
>>> > >> > > So, I
>>> > >> > > > > > think
>>> > >> > > > > > > > we
>>> > >> > > > > > > > > > don't care about this case. (I'm sorry, I am not
>>> > >> familiar
>>> > >> > > with
>>> > >> > > > > > > > > down-convert
>>> > >> > > > > > > > > > logic.)
>>> > >> > > > > > > > > >
>>> > >> > > > > > > > > > Please give more opinions. Thanks!
>>> > >> > > > > > > > > >
>>> > >> > > > > > > > > > - Dongjin
>>> > >> > > > > > > > > >
>>> > >> > > > > > > > > >
>>> > >> > > > > > > > > > On Wed, Aug 8, 2018 at 6:41 AM Jason Gustafson <
>>> > >> > > > > jason@confluent.io
>>> > >> > > > > > >
>>> > >> > > > > > > > > wrote:
>>> > >> > > > > > > > > >
>>> > >> > > > > > > > > > > Hey Colin,
>>> > >> > > > > > > > > > >
>>> > >> > > > > > > > > > > The problem for the fetch API is that the broker
>>> > does
>>> > >> not
>>> > >> > > > > > generally
>>> > >> > > > > > > > > know
>>> > >> > > > > > > > > > if
>>> > >> > > > > > > > > > > a batch was compressed with zstd unless it
>>> parses
>>> > it.
>>> > >> I
>>> > >> > > think
>>> > >> > > > > the
>>> > >> > > > > > > > goal
>>> > >> > > > > > > > > > here
>>> > >> > > > > > > > > > > is to avoid the expensive down-conversion that
>>> is
>>> > >> needed
>>> > >> > to
>>> > >> > > > > > ensure
>>> > >> > > > > > > > > > > compatibility because it is only necessary if
>>> zstd
>>> > is
>>> > >> > > > actually
>>> > >> > > > > in
>>> > >> > > > > > > > use.
>>> > >> > > > > > > > > > But
>>> > >> > > > > > > > > > > as long as old clients can parse the message
>>> format,
>>> > >> they
>>> > >> > > > > should
>>> > >> > > > > > > get
>>> > >> > > > > > > > a
>>> > >> > > > > > > > > > > reasonable error if they see an unsupported
>>> > >> compression
>>> > >> > > type
>>> > >> > > > in
>>> > >> > > > > > the
>>> > >> > > > > > > > > > > attributes. Basically the onus is on users to
>>> ensure
>>> > >> that
>>> > >> > > > their
>>> > >> > > > > > > > > consumers
>>> > >> > > > > > > > > > > have been updated prior to using zstd. It seems
>>> > like a
>>> > >> > > > > reasonable
>>> > >> > > > > > > > > > tradeoff
>>> > >> > > > > > > > > > > to me. There are a couple cases that might be
>>> worth
>>> > >> > > thinking
>>> > >> > > > > > > through:
>>> > >> > > > > > > > > > >
>>> > >> > > > > > > > > > > 1. If a topic is explicitly configured to use
>>> zstd,
>>> > >> then
>>> > >> > we
>>> > >> > > > > don't
>>> > >> > > > > > > > need
>>> > >> > > > > > > > > to
>>> > >> > > > > > > > > > > check the fetched data for the compression type
>>> to
>>> > >> know
>>> > >> > if
>>> > >> > > we
>>> > >> > > > > > need
>>> > >> > > > > > > > > > > down-conversion. If we did bump the Fetch API
>>> > version,
>>> > >> > then
>>> > >> > > > we
>>> > >> > > > > > > could
>>> > >> > > > > > > > > > handle
>>> > >> > > > > > > > > > > this case by either down-converting using a
>>> > different
>>> > >> > > > > compression
>>> > >> > > > > > > > type
>>> > >> > > > > > > > > or
>>> > >> > > > > > > > > > > returning an error.
>>> > >> > > > > > > > > > > 2. Similarly, if we have to down-convert anyway
>>> > >> because
>>> > >> > the
>>> > >> > > > > > client
>>> > >> > > > > > > > does
>>> > >> > > > > > > > > > not
>>> > >> > > > > > > > > > > understand the message format, then we could
>>> also
>>> > use
>>> > >> a
>>> > >> > > > > different
>>> > >> > > > > > > > > > > compression type.
>>> > >> > > > > > > > > > >
>>> > >> > > > > > > > > > > For the produce API, I think it's reasonable to
>>> bump
>>> > >> the
>>> > >> > > api
>>> > >> > > > > > > version.
>>> > >> > > > > > > > > > This
>>> > >> > > > > > > > > > > can be used by clients to check whether a broker
>>> > >> supports
>>> > >> > > > zstd.
>>> > >> > > > > > For
>>> > >> > > > > > > > > > > example, we might support a list of preferred
>>> > >> compression
>>> > >> > > > types
>>> > >> > > > > > in
>>> > >> > > > > > > > the
>>> > >> > > > > > > > > > > producer and we could use the broker to detect
>>> which
>>> > >> > > version
>>> > >> > > > to
>>> > >> > > > > > > use.
>>> > >> > > > > > > > > > >
>>> > >> > > > > > > > > > > -Jason
>>> > >> > > > > > > > > > >
>>> > >> > > > > > > > > > > On Tue, Aug 7, 2018 at 1:32 PM, Colin McCabe <
>>> > >> > > > > cmccabe@apache.org
>>> > >> > > > > > >
>>> > >> > > > > > > > > wrote:
>>> > >> > > > > > > > > > >
>>> > >> > > > > > > > > > > > Thanks for bumping this, Dongjin.  ZStd is a
>>> good
>>> > >> > > > compression
>>> > >> > > > > > > codec
>>> > >> > > > > > > > > > and I
>>> > >> > > > > > > > > > > > hope we can get this support in soon!
>>> > >> > > > > > > > > > > >
>>> > >> > > > > > > > > > > > I would say we can just bump the API version
>>> to
>>> > >> > indicate
>>> > >> > > > that
>>> > >> > > > > > > ZStd
>>> > >> > > > > > > > > > > support
>>> > >> > > > > > > > > > > > is expected in new clients.  We probably need
>>> some
>>> > >> way
>>> > >> > of
>>> > >> > > > > > > > indicating
>>> > >> > > > > > > > > to
>>> > >> > > > > > > > > > > the
>>> > >> > > > > > > > > > > > older clients that they can't consume the
>>> > >> partitions,
>>> > >> > as
>>> > >> > > > > well.
>>> > >> > > > > > > > > Perhaps
>>> > >> > > > > > > > > > > we
>>> > >> > > > > > > > > > > > can use the UNSUPPORTED_FOR_MESSAGE_FORMAT
>>> error?
>>> > >> > > > > > > > > > > >
>>> > >> > > > > > > > > > > > The license thing seems straightforward --
>>> it's
>>> > >> just a
>>> > >> > > > matter
>>> > >> > > > > > of
>>> > >> > > > > > > > > adding
>>> > >> > > > > > > > > > > > the text to the right files as per ASF
>>> guidelines.
>>> > >> > > > > > > > > > > >
>>> > >> > > > > > > > > > > > With regard to the dictionary support, do we
>>> > really
>>> > >> > need
>>> > >> > > > that
>>> > >> > > > > > in
>>> > >> > > > > > > > the
>>> > >> > > > > > > > > > > first
>>> > >> > > > > > > > > > > > version?  Hopefully message batches are big
>>> enough
>>> > >> that
>>> > >> > > > this
>>> > >> > > > > > > isn't
>>> > >> > > > > > > > > > > needed.
>>> > >> > > > > > > > > > > > Some benchmarks might help here.
>>> > >> > > > > > > > > > > >
>>> > >> > > > > > > > > > > > best,
>>> > >> > > > > > > > > > > > Colin
>>> > >> > > > > > > > > > > >
>>> > >> > > > > > > > > > > >
>>> > >> > > > > > > > > > > > On Tue, Aug 7, 2018, at 08:02, Dongjin Lee
>>> wrote:
>>> > >> > > > > > > > > > > > > As Kafka 2.0.0 was released, let's reboot
>>> this
>>> > >> issue,
>>> > >> > > > > KIP-110
>>> > >> > > > > > > > > > > > > <https://cwiki.apache.org/
>>> > >> > > confluence/display/KAFKA/KIP-
>>> > >> > > > > > > > > > > > 110%3A+Add+Codec+for+ZStandard+Compression>
>>> > >> > > > > > > > > > > > > .
>>> > >> > > > > > > > > > > > >
>>> > >> > > > > > > > > > > > > For newcomers, Here is some summary of the
>>> > >> history:
>>> > >> > > > KIP-110
>>> > >> > > > > > was
>>> > >> > > > > > > > > > > > originally
>>> > >> > > > > > > > > > > > > worked for the issue KAFKA-4514 but, it
>>> lacked
>>> > >> > > benchmark
>>> > >> > > > > > > results
>>> > >> > > > > > > > to
>>> > >> > > > > > > > > > get
>>> > >> > > > > > > > > > > > the
>>> > >> > > > > > > > > > > > > agreement of the community. Later, Ivan
>>> Babrou
>>> > and
>>> > >> > some
>>> > >> > > > > other
>>> > >> > > > > > > > users
>>> > >> > > > > > > > > > who
>>> > >> > > > > > > > > > > > > adopted the patch provided their excellent
>>> > >> > performance
>>> > >> > > > > report
>>> > >> > > > > > > > which
>>> > >> > > > > > > > > > is
>>> > >> > > > > > > > > > > > now
>>> > >> > > > > > > > > > > > > included in the KIP, but it postponed again
>>> > >> because
>>> > >> > of
>>> > >> > > > the
>>> > >> > > > > > > > > community
>>> > >> > > > > > > > > > > was
>>> > >> > > > > > > > > > > > > busy for 2.0.0 release. It is why I now
>>> reboot
>>> > >> this
>>> > >> > > > issue.
>>> > >> > > > > > > > > > > > >
>>> > >> > > > > > > > > > > > > The following is the current status of the
>>> > >> feature:
>>> > >> > You
>>> > >> > > > can
>>> > >> > > > > > > check
>>> > >> > > > > > > > > the
>>> > >> > > > > > > > > > > > > current draft implementation here
>>> > >> > > > > > > > > > > > > <https://github.com/apache/kafka/pull/2267>.
>>> It
>>> > >> is
>>> > >> > > based
>>> > >> > > > > on
>>> > >> > > > > > > zstd
>>> > >> > > > > > > > > > 1.3.5
>>> > >> > > > > > > > > > > > and
>>> > >> > > > > > > > > > > > > periodically rebased onto the latest
>>> trunk[^1].
>>> > >> > > > > > > > > > > > >
>>> > >> > > > > > > > > > > > > The issues that should be addressed is like
>>> > >> > following:
>>> > >> > > > > > > > > > > > >
>>> > >> > > > > > > > > > > > > *1. Backward Compatibility*
>>> > >> > > > > > > > > > > > >
>>> > >> > > > > > > > > > > > > To support old consumers, we need to take a
>>> > >> strategy
>>> > >> > to
>>> > >> > > > > > handle
>>> > >> > > > > > > > the
>>> > >> > > > > > > > > > old
>>> > >> > > > > > > > > > > > > consumers. Current candidates are:
>>> > >> > > > > > > > > > > > >
>>> > >> > > > > > > > > > > > > - Bump API version
>>> > >> > > > > > > > > > > > > - Leave unchanged: let the old clients fail.
>>> > >> > > > > > > > > > > > > - Improve the error messages:
>>> > >> > > > > > > > > > > > >
>>> > >> > > > > > > > > > > > > *2. Dictionary Support*
>>> > >> > > > > > > > > > > > >
>>> > >> > > > > > > > > > > > > To support zstd's dictionary feature in the
>>> > future
>>> > >> > (if
>>> > >> > > > > > needed),
>>> > >> > > > > > > > we
>>> > >> > > > > > > > > > need
>>> > >> > > > > > > > > > > > to
>>> > >> > > > > > > > > > > > > sketch how it should be and leave some room
>>> for
>>> > >> it.
>>> > >> > As
>>> > >> > > of
>>> > >> > > > > > now,
>>> > >> > > > > > > > > there
>>> > >> > > > > > > > > > > has
>>> > >> > > > > > > > > > > > > been no discussion on this topic yet.
>>> > >> > > > > > > > > > > > >
>>> > >> > > > > > > > > > > > > *3. License*
>>> > >> > > > > > > > > > > > >
>>> > >> > > > > > > > > > > > > To use this feature, we need to add license
>>> of
>>> > >> zstd
>>> > >> > and
>>> > >> > > > > > > zstd-jni
>>> > >> > > > > > > > to
>>> > >> > > > > > > > > > the
>>> > >> > > > > > > > > > > > > project. (Thanks to Viktor Somogyi for
>>> raising
>>> > >> this
>>> > >> > > > issue!)
>>> > >> > > > > > It
>>> > >> > > > > > > > > seems
>>> > >> > > > > > > > > > > like
>>> > >> > > > > > > > > > > > > what Apache Spark did would be a good
>>> example
>>> > but
>>> > >> > there
>>> > >> > > > has
>>> > >> > > > > > > been
>>> > >> > > > > > > > no
>>> > >> > > > > > > > > > > > > discussion yet.
>>> > >> > > > > > > > > > > > >
>>> > >> > > > > > > > > > > > > You can find the details of the above
>>> issues in
>>> > >> the
>>> > >> > KIP
>>> > >> > > > > > > document.
>>> > >> > > > > > > > > > > Please
>>> > >> > > > > > > > > > > > > have a look when you are free, and give me
>>> > >> feedback.
>>> > >> > > All
>>> > >> > > > > > kinds
>>> > >> > > > > > > of
>>> > >> > > > > > > > > > > > > participating are welcome.
>>> > >> > > > > > > > > > > > >
>>> > >> > > > > > > > > > > > > Best,
>>> > >> > > > > > > > > > > > > Dongjin
>>> > >> > > > > > > > > > > > >
>>> > >> > > > > > > > > > > > > [^1]: At the time of writing, commit
>>> 6b4fb8152.
>>> > >> > > > > > > > > > > > >
>>> > >> > > > > > > > > > > > > On Sat, Jul 14, 2018 at 10:45 PM Dongjin
>>> Lee <
>>> > >> > > > > > > dongjin@apache.org
>>> > >> > > > > > > > >
>>> > >> > > > > > > > > > > wrote:
>>> > >> > > > > > > > > > > > >
>>> > >> > > > > > > > > > > > > > Sorry for the late reply.
>>> > >> > > > > > > > > > > > > >
>>> > >> > > > > > > > > > > > > > In short, I could not submit the updated
>>> KIP
>>> > by
>>> > >> the
>>> > >> > > > > feature
>>> > >> > > > > > > > > freeze
>>> > >> > > > > > > > > > > > > > deadline of 2.0.0. For this reason, it
>>> will
>>> > not
>>> > >> be
>>> > >> > > > > included
>>> > >> > > > > > > in
>>> > >> > > > > > > > > the
>>> > >> > > > > > > > > > > > 2.0.0
>>> > >> > > > > > > > > > > > > > release and all discussion for this issue
>>> were
>>> > >> > > > postponed
>>> > >> > > > > > > after
>>> > >> > > > > > > > > the
>>> > >> > > > > > > > > > > > release
>>> > >> > > > > > > > > > > > > > of 2.0.0.
>>> > >> > > > > > > > > > > > > >
>>> > >> > > > > > > > > > > > > > I have been updating the PR following
>>> recent
>>> > >> > updates.
>>> > >> > > > > Just
>>> > >> > > > > > > > now, I
>>> > >> > > > > > > > > > > > rebased
>>> > >> > > > > > > > > > > > > > it against the latest trunk and updated
>>> the
>>> > zstd
>>> > >> > > > version
>>> > >> > > > > > into
>>> > >> > > > > > > > > > 1.3.5.
>>> > >> > > > > > > > > > > > If you
>>> > >> > > > > > > > > > > > > > need some request, don't hesitate to
>>> notify
>>> > me.
>>> > >> > (But
>>> > >> > > > not
>>> > >> > > > > > this
>>> > >> > > > > > > > > > thread
>>> > >> > > > > > > > > > > -
>>> > >> > > > > > > > > > > > just
>>> > >> > > > > > > > > > > > > > send me the message directly.)
>>> > >> > > > > > > > > > > > > >
>>> > >> > > > > > > > > > > > > > Best,
>>> > >> > > > > > > > > > > > > > Dongjin
>>> > >> > > > > > > > > > > > > >
>>> > >> > > > > > > > > > > > > > On Tue, Jul 10, 2018 at 11:57 PM Bobby
>>> Evans <
>>> > >> > > > > > > bobby@apache.org
>>> > >> > > > > > > > >
>>> > >> > > > > > > > > > > wrote:
>>> > >> > > > > > > > > > > > > >
>>> > >> > > > > > > > > > > > > >> I there any update on this.  The
>>> performance
>>> > >> > > > > improvements
>>> > >> > > > > > > are
>>> > >> > > > > > > > > > quite
>>> > >> > > > > > > > > > > > > >> impressive and I really would like to
>>> stop
>>> > >> forking
>>> > >> > > > kafka
>>> > >> > > > > > > just
>>> > >> > > > > > > > to
>>> > >> > > > > > > > > > get
>>> > >> > > > > > > > > > > > this
>>> > >> > > > > > > > > > > > > >> in.
>>> > >> > > > > > > > > > > > > >>
>>> > >> > > > > > > > > > > > > >> Thanks,
>>> > >> > > > > > > > > > > > > >>
>>> > >> > > > > > > > > > > > > >> Bobby
>>> > >> > > > > > > > > > > > > >>
>>> > >> > > > > > > > > > > > > >> On Wed, Jun 13, 2018 at 8:56 PM Dongjin
>>> Lee <
>>> > >> > > > > > > > dongjin@apache.org
>>> > >> > > > > > > > > >
>>> > >> > > > > > > > > > > > wrote:
>>> > >> > > > > > > > > > > > > >>
>>> > >> > > > > > > > > > > > > >> > Ismael,
>>> > >> > > > > > > > > > > > > >> >
>>> > >> > > > > > > > > > > > > >> > Oh, I forgot all of you are on working
>>> > frenzy
>>> > >> > for
>>> > >> > > > 2.0!
>>> > >> > > > > > No
>>> > >> > > > > > > > > > problem,
>>> > >> > > > > > > > > > > > take
>>> > >> > > > > > > > > > > > > >> > your time. I am also working at another
>>> > issue
>>> > >> > now.
>>> > >> > > > > Thank
>>> > >> > > > > > > you
>>> > >> > > > > > > > > for
>>> > >> > > > > > > > > > > > > >> letting me
>>> > >> > > > > > > > > > > > > >> > know.
>>> > >> > > > > > > > > > > > > >> >
>>> > >> > > > > > > > > > > > > >> > Best,
>>> > >> > > > > > > > > > > > > >> > Dongjin
>>> > >> > > > > > > > > > > > > >> >
>>> > >> > > > > > > > > > > > > >> > On Wed, Jun 13, 2018, 11:44 PM Ismael
>>> Juma
>>> > <
>>> > >> > > > > > > > ismaelj@gmail.com
>>> > >> > > > > > > > > >
>>> > >> > > > > > > > > > > > wrote:
>>> > >> > > > > > > > > > > > > >> >
>>> > >> > > > > > > > > > > > > >> > > Sorry for the delay Dongjin.
>>> Everyone is
>>> > >> busy
>>> > >> > > > > > finalising
>>> > >> > > > > > > > > > 2.0.0.
>>> > >> > > > > > > > > > > > This
>>> > >> > > > > > > > > > > > > >> KIP
>>> > >> > > > > > > > > > > > > >> > > seems like a great candidate for
>>> 2.1.0
>>> > and
>>> > >> > > > hopefully
>>> > >> > > > > > > there
>>> > >> > > > > > > > > > will
>>> > >> > > > > > > > > > > be
>>> > >> > > > > > > > > > > > > >> more
>>> > >> > > > > > > > > > > > > >> > of
>>> > >> > > > > > > > > > > > > >> > > a discussion next week. :)
>>> > >> > > > > > > > > > > > > >> > >
>>> > >> > > > > > > > > > > > > >> > > Ismael
>>> > >> > > > > > > > > > > > > >> > >
>>> > >> > > > > > > > > > > > > >> > > On Wed, 13 Jun 2018, 05:17 Dongjin
>>> Lee, <
>>> > >> > > > > > > > dongjin@apache.org
>>> > >> > > > > > > > > >
>>> > >> > > > > > > > > > > > wrote:
>>> > >> > > > > > > > > > > > > >> > >
>>> > >> > > > > > > > > > > > > >> > > > Hello. I just updated my draft
>>> > >> > implementation:
>>> > >> > > > > > > > > > > > > >> > > >
>>> > >> > > > > > > > > > > > > >> > > > 1. Rebased to latest trunk (commit
>>> > >> 5145d6b)
>>> > >> > > > > > > > > > > > > >> > > > 2. Apply ZStd 1.3.4
>>> > >> > > > > > > > > > > > > >> > > >
>>> > >> > > > > > > > > > > > > >> > > > You can check out the
>>> implementation
>>> > from
>>> > >> > here
>>> > >> > > > > > > > > > > > > >> > > > <
>>> > >> https://github.com/apache/kafka/pull/2267
>>> > >> > >.
>>> > >> > > If
>>> > >> > > > > you
>>> > >> > > > > > > > > > > experience
>>> > >> > > > > > > > > > > > any
>>> > >> > > > > > > > > > > > > >> > > problem
>>> > >> > > > > > > > > > > > > >> > > > running it, don't hesitate to give
>>> me a
>>> > >> > > mention.
>>> > >> > > > > > > > > > > > > >> > > >
>>> > >> > > > > > > > > > > > > >> > > > Best,
>>> > >> > > > > > > > > > > > > >> > > > Dongjin
>>> > >> > > > > > > > > > > > > >> > > >
>>> > >> > > > > > > > > > > > > >> > > > On Tue, Jun 12, 2018 at 6:50 PM
>>> Dongjin
>>> > >> Lee
>>> > >> > <
>>> > >> > > > > > > > > > > dongjin@apache.org
>>> > >> > > > > > > > > > > > >
>>> > >> > > > > > > > > > > > > >> > wrote:
>>> > >> > > > > > > > > > > > > >> > > >
>>> > >> > > > > > > > > > > > > >> > > > > Here is the short conclusion
>>> about
>>> > the
>>> > >> > > license
>>> > >> > > > > > > > problem:
>>> > >> > > > > > > > > > *We
>>> > >> > > > > > > > > > > > can
>>> > >> > > > > > > > > > > > > >> use
>>> > >> > > > > > > > > > > > > >> > > zstd
>>> > >> > > > > > > > > > > > > >> > > > > and zstd-jni without any
>>> problem, but
>>> > >> we
>>> > >> > > need
>>> > >> > > > to
>>> > >> > > > > > > > include
>>> > >> > > > > > > > > > > their
>>> > >> > > > > > > > > > > > > >> > license,
>>> > >> > > > > > > > > > > > > >> > > > > e.g., BSD license.*
>>> > >> > > > > > > > > > > > > >> > > > >
>>> > >> > > > > > > > > > > > > >> > > > > Both of BSD 2 Clause License & 3
>>> > Clause
>>> > >> > > > License
>>> > >> > > > > > > > requires
>>> > >> > > > > > > > > > to
>>> > >> > > > > > > > > > > > > >> include
>>> > >> > > > > > > > > > > > > >> > the
>>> > >> > > > > > > > > > > > > >> > > > > license used, and BSD 3 Clause
>>> > License
>>> > >> > > > requires
>>> > >> > > > > > that
>>> > >> > > > > > > > the
>>> > >> > > > > > > > > > > name
>>> > >> > > > > > > > > > > > of
>>> > >> > > > > > > > > > > > > >> the
>>> > >> > > > > > > > > > > > > >> > > > > contributor can't be used to
>>> endorse
>>> > or
>>> > >> > > > promote
>>> > >> > > > > > the
>>> > >> > > > > > > > > > product.
>>> > >> > > > > > > > > > > > > >> That's
>>> > >> > > > > > > > > > > > > >> > it
>>> > >> > > > > > > > > > > > > >> > > > > <
>>> > >> > > > > > > > > > > > > >> > > >
>>> > >> > > > > > > > > > > > > >> > >
>>> > >> > > > > > > > > > > > > >> >
>>> > >> > > > > > > > > > > > > >>
>>> > >> > http://www.mikestratton.net/2011/12/is-bsd-license-
>>> > >> > > > > > > > > > > > compatible-with-apache-2-0-license/
>>> > >> > > > > > > > > > > > > >> > > > >
>>> > >> > > > > > > > > > > > > >> > > > > - They are not listed in the
>>> list of
>>> > >> > > > prohibited
>>> > >> > > > > > > > licenses
>>> > >> > > > > > > > > > > > > >> > > > > <https://www.apache.org/legal/
>>> > >> > > > > > > > resolved.html#category-x>
>>> > >> > > > > > > > > > > also.
>>> > >> > > > > > > > > > > > > >> > > > >
>>> > >> > > > > > > > > > > > > >> > > > > Here is how Spark did for it
>>> > >> > > > > > > > > > > > > >> > > > > <https://issues.apache.org/
>>> > >> > > > > > jira/browse/SPARK-19112
>>> > >> > > > > > > >:
>>> > >> > > > > > > > > > > > > >> > > > >
>>> > >> > > > > > > > > > > > > >> > > > > - They made a directory
>>> dedicated to
>>> > >> the
>>> > >> > > > > > dependency
>>> > >> > > > > > > > > > license
>>> > >> > > > > > > > > > > > files
>>> > >> > > > > > > > > > > > > >> > > > > <
>>> > >> > > > > > > https://github.com/apache/spark/tree/master/licenses
>>> > >> > > > > > > > >
>>> > >> > > > > > > > > > and
>>> > >> > > > > > > > > > > > added
>>> > >> > > > > > > > > > > > > >> > > > licenses
>>> > >> > > > > > > > > > > > > >> > > > > for Zstd
>>> > >> > > > > > > > > > > > > >> > > > > <
>>> > >> > > > > > > > > > > > > >> >
>>> > >> > > > https://github.com/apache/spark/blob/master/licenses/
>>> > >> > > > > > > > > > > > LICENSE-zstd.txt
>>> > >> > > > > > > > > > > > > >> > > >
>>> > >> > > > > > > > > > > > > >> > > > &
>>> > >> > > > > > > > > > > > > >> > > > > Zstd-jni
>>> > >> > > > > > > > > > > > > >> > > > > <
>>> > >> > > > > > > > > > > > > >> > > >
>>> > >> > > > > > > > > > > > > >> > >
>>> > >> > > > > > > > > > > > > >> >
>>> > >> > > > > > > > > > > > > >> https://github.com/apache/
>>> > >> > > spark/blob/master/licenses/
>>> > >> > > > > > > > > > > > LICENSE-zstd-jni.txt
>>> > >> > > > > > > > > > > > > >> >
>>> > >> > > > > > > > > > > > > >> > > > > .
>>> > >> > > > > > > > > > > > > >> > > > > - Added a link to the original
>>> > license
>>> > >> > files
>>> > >> > > > in
>>> > >> > > > > > > > LICENSE.
>>> > >> > > > > > > > > > > > > >> > > > > <
>>> > >> > > > > https://github.com/apache/spark/pull/18805/files
>>> > >> > > > > > >
>>> > >> > > > > > > > > > > > > >> > > > >
>>> > >> > > > > > > > > > > > > >> > > > > If needed, I can make a similar
>>> > update.
>>> > >> > > > > > > > > > > > > >> > > > >
>>> > >> > > > > > > > > > > > > >> > > > > Thanks for pointing out this
>>> problem,
>>> > >> > > Viktor!
>>> > >> > > > > Nice
>>> > >> > > > > > > > > catch!
>>> > >> > > > > > > > > > > > > >> > > > >
>>> > >> > > > > > > > > > > > > >> > > > > Best,
>>> > >> > > > > > > > > > > > > >> > > > > Dongjin
>>> > >> > > > > > > > > > > > > >> > > > >
>>> > >> > > > > > > > > > > > > >> > > > >
>>> > >> > > > > > > > > > > > > >> > > > >
>>> > >> > > > > > > > > > > > > >> > > > > On Mon, Jun 11, 2018 at 11:50 PM
>>> > >> Dongjin
>>> > >> > > Lee <
>>> > >> > > > > > > > > > > > dongjin@apache.org>
>>> > >> > > > > > > > > > > > > >> > > wrote:
>>> > >> > > > > > > > > > > > > >> > > > >
>>> > >> > > > > > > > > > > > > >> > > > >> I greatly appreciate your
>>> > >> comprehensive
>>> > >> > > > > > reasoning.
>>> > >> > > > > > > > so:
>>> > >> > > > > > > > > +1
>>> > >> > > > > > > > > > > > for b
>>> > >> > > > > > > > > > > > > >> > until
>>> > >> > > > > > > > > > > > > >> > > > now.
>>> > >> > > > > > > > > > > > > >> > > > >>
>>> > >> > > > > > > > > > > > > >> > > > >> For the license issues, I will
>>> have
>>> > a
>>> > >> > check
>>> > >> > > > on
>>> > >> > > > > > how
>>> > >> > > > > > > > the
>>> > >> > > > > > > > > > over
>>> > >> > > > > > > > > > > > > >> projects
>>> > >> > > > > > > > > > > > > >> > > are
>>> > >> > > > > > > > > > > > > >> > > > >> doing and share the results.
>>> > >> > > > > > > > > > > > > >> > > > >>
>>> > >> > > > > > > > > > > > > >> > > > >> Best,
>>> > >> > > > > > > > > > > > > >> > > > >> Dongjin
>>> > >> > > > > > > > > > > > > >> > > > >>
>>> > >> > > > > > > > > > > > > >> > > > >> On Mon, Jun 11, 2018 at 10:08 PM
>>> > >> Viktor
>>> > >> > > > > Somogyi <
>>> > >> > > > > > > > > > > > > >> > > > viktorsomogyi@gmail.com>
>>> > >> > > > > > > > > > > > > >> > > > >> wrote:
>>> > >> > > > > > > > > > > > > >> > > > >>
>>> > >> > > > > > > > > > > > > >> > > > >>> Hi Dongjin,
>>> > >> > > > > > > > > > > > > >> > > > >>>
>>> > >> > > > > > > > > > > > > >> > > > >>> A couple of comments:
>>> > >> > > > > > > > > > > > > >> > > > >>> I would vote for option b. in
>>> the
>>> > >> > > "backward
>>> > >> > > > > > > > > > compatibility"
>>> > >> > > > > > > > > > > > > >> section.
>>> > >> > > > > > > > > > > > > >> > > My
>>> > >> > > > > > > > > > > > > >> > > > >>> reasoning for this is that
>>> users
>>> > >> > upgrading
>>> > >> > > > to
>>> > >> > > > > a
>>> > >> > > > > > > zstd
>>> > >> > > > > > > > > > > > compatible
>>> > >> > > > > > > > > > > > > >> > > version
>>> > >> > > > > > > > > > > > > >> > > > >>> won't start to use it
>>> > automatically,
>>> > >> so
>>> > >> > > > manual
>>> > >> > > > > > > > > > > > reconfiguration
>>> > >> > > > > > > > > > > > > >> is
>>> > >> > > > > > > > > > > > > >> > > > >>> required.
>>> > >> > > > > > > > > > > > > >> > > > >>> Therefore an upgrade won't
>>> mess up
>>> > >> the
>>> > >> > > > > cluster.
>>> > >> > > > > > If
>>> > >> > > > > > > > not
>>> > >> > > > > > > > > > all
>>> > >> > > > > > > > > > > > the
>>> > >> > > > > > > > > > > > > >> > > clients
>>> > >> > > > > > > > > > > > > >> > > > >>> are
>>> > >> > > > > > > > > > > > > >> > > > >>> upgraded but just some of them
>>> and
>>> > >> > they'd
>>> > >> > > > > start
>>> > >> > > > > > to
>>> > >> > > > > > > > use
>>> > >> > > > > > > > > > > zstd
>>> > >> > > > > > > > > > > > > >> then it
>>> > >> > > > > > > > > > > > > >> > > > would
>>> > >> > > > > > > > > > > > > >> > > > >>> cause errors in the cluster.
>>> I'd
>>> > >> like to
>>> > >> > > > > presume
>>> > >> > > > > > > > > though
>>> > >> > > > > > > > > > > that
>>> > >> > > > > > > > > > > > > >> this
>>> > >> > > > > > > > > > > > > >> > is
>>> > >> > > > > > > > > > > > > >> > > a
>>> > >> > > > > > > > > > > > > >> > > > >>> very
>>> > >> > > > > > > > > > > > > >> > > > >>> obvious failure case and nobody
>>> > >> should
>>> > >> > be
>>> > >> > > > > > > surprised
>>> > >> > > > > > > > if
>>> > >> > > > > > > > > > it
>>> > >> > > > > > > > > > > > didn't
>>> > >> > > > > > > > > > > > > >> > > work.
>>> > >> > > > > > > > > > > > > >> > > > >>> I wouldn't choose a. as I
>>> think we
>>> > >> > should
>>> > >> > > > bump
>>> > >> > > > > > the
>>> > >> > > > > > > > > fetch
>>> > >> > > > > > > > > > > and
>>> > >> > > > > > > > > > > > > >> > produce
>>> > >> > > > > > > > > > > > > >> > > > >>> requests if it's a change in
>>> the
>>> > >> message
>>> > >> > > > > format.
>>> > >> > > > > > > > > > Moreover
>>> > >> > > > > > > > > > > if
>>> > >> > > > > > > > > > > > > >> some
>>> > >> > > > > > > > > > > > > >> > of
>>> > >> > > > > > > > > > > > > >> > > > the
>>> > >> > > > > > > > > > > > > >> > > > >>> producers and the brokers are
>>> > >> upgraded
>>> > >> > but
>>> > >> > > > > some
>>> > >> > > > > > of
>>> > >> > > > > > > > the
>>> > >> > > > > > > > > > > > consumers
>>> > >> > > > > > > > > > > > > >> > are
>>> > >> > > > > > > > > > > > > >> > > > not,
>>> > >> > > > > > > > > > > > > >> > > > >>> then we wouldn't prevent the
>>> error
>>> > >> when
>>> > >> > > the
>>> > >> > > > > old
>>> > >> > > > > > > > > consumer
>>> > >> > > > > > > > > > > > tries
>>> > >> > > > > > > > > > > > > >> to
>>> > >> > > > > > > > > > > > > >> > > > consume
>>> > >> > > > > > > > > > > > > >> > > > >>> the zstd compressed messages.
>>> > >> > > > > > > > > > > > > >> > > > >>> I wouldn't choose c. either as
>>> I
>>> > >> think
>>> > >> > > > binding
>>> > >> > > > > > the
>>> > >> > > > > > > > > > > > compression
>>> > >> > > > > > > > > > > > > >> type
>>> > >> > > > > > > > > > > > > >> > > to
>>> > >> > > > > > > > > > > > > >> > > > an
>>> > >> > > > > > > > > > > > > >> > > > >>> API is not so obvious from the
>>> > >> > developer's
>>> > >> > > > > > > > > perspective.
>>> > >> > > > > > > > > > > > > >> > > > >>>
>>> > >> > > > > > > > > > > > > >> > > > >>> I would also prefer to use the
>>> > >> existing
>>> > >> > > > > binding,
>>> > >> > > > > > > > > however
>>> > >> > > > > > > > > > > we
>>> > >> > > > > > > > > > > > must
>>> > >> > > > > > > > > > > > > >> > > > respect
>>> > >> > > > > > > > > > > > > >> > > > >>> the licenses:
>>> > >> > > > > > > > > > > > > >> > > > >>> "The code for these JNI
>>> bindings is
>>> > >> > > licenced
>>> > >> > > > > > under
>>> > >> > > > > > > > > > > 2-clause
>>> > >> > > > > > > > > > > > BSD
>>> > >> > > > > > > > > > > > > >> > > > license.
>>> > >> > > > > > > > > > > > > >> > > > >>> The native Zstd library is
>>> licensed
>>> > >> > under
>>> > >> > > > > > 3-clause
>>> > >> > > > > > > > BSD
>>> > >> > > > > > > > > > > > license
>>> > >> > > > > > > > > > > > > >> and
>>> > >> > > > > > > > > > > > > >> > > > GPL2"
>>> > >> > > > > > > > > > > > > >> > > > >>> Based on the FAQ page
>>> > >> > > > > > > > > > > > > >> > > > >>> https://www.apache.org/legal/
>>> > >> > > > > > > > resolved.html#category-a
>>> > >> > > > > > > > > > > > > >> > > > >>> we may use 2- and 3-clause BSD
>>> > >> licenses
>>> > >> > > but
>>> > >> > > > > the
>>> > >> > > > > > > > Apache
>>> > >> > > > > > > > > > > > license
>>> > >> > > > > > > > > > > > > >> is
>>> > >> > > > > > > > > > > > > >> > not
>>> > >> > > > > > > > > > > > > >> > > > >>> compatible with GPL2. I'm
>>> hoping
>>> > that
>>> > >> > the
>>> > >> > > > > > > "3-clause
>>> > >> > > > > > > > > BSD
>>> > >> > > > > > > > > > > > license
>>> > >> > > > > > > > > > > > > >> and
>>> > >> > > > > > > > > > > > > >> > > > GPL2"
>>> > >> > > > > > > > > > > > > >> > > > >>> is really not an AND but an OR
>>> in
>>> > >> this
>>> > >> > > case,
>>> > >> > > > > but
>>> > >> > > > > > > I'm
>>> > >> > > > > > > > > no
>>> > >> > > > > > > > > > > > lawyer,
>>> > >> > > > > > > > > > > > > >> > just
>>> > >> > > > > > > > > > > > > >> > > > >>> wanted
>>> > >> > > > > > > > > > > > > >> > > > >>> to make the point that we
>>> should
>>> > >> watch
>>> > >> > out
>>> > >> > > > for
>>> > >> > > > > > > > > licenses.
>>> > >> > > > > > > > > > > :)
>>> > >> > > > > > > > > > > > > >> > > > >>>
>>> > >> > > > > > > > > > > > > >> > > > >>> Regards,
>>> > >> > > > > > > > > > > > > >> > > > >>> Viktor
>>> > >> > > > > > > > > > > > > >> > > > >>>
>>> > >> > > > > > > > > > > > > >> > > > >>>
>>> > >> > > > > > > > > > > > > >> > > > >>> On Sun, Jun 10, 2018 at 3:02 AM
>>> > Ivan
>>> > >> > > Babrou
>>> > >> > > > <
>>> > >> > > > > > > > > > > > ibobrik@gmail.com>
>>> > >> > > > > > > > > > > > > >> > > wrote:
>>> > >> > > > > > > > > > > > > >> > > > >>>
>>> > >> > > > > > > > > > > > > >> > > > >>> > Hello,
>>> > >> > > > > > > > > > > > > >> > > > >>> >
>>> > >> > > > > > > > > > > > > >> > > > >>> > This is Ivan and I still very
>>> > much
>>> > >> > > support
>>> > >> > > > > the
>>> > >> > > > > > > > fact
>>> > >> > > > > > > > > > that
>>> > >> > > > > > > > > > > > zstd
>>> > >> > > > > > > > > > > > > >> > > > >>> compression
>>> > >> > > > > > > > > > > > > >> > > > >>> > should be included out of the
>>> > box.
>>> > >> > > > > > > > > > > > > >> > > > >>> >
>>> > >> > > > > > > > > > > > > >> > > > >>> > Please think about the
>>> > environment,
>>> > >> > you
>>> > >> > > > can
>>> > >> > > > > > save
>>> > >> > > > > > > > > > quite a
>>> > >> > > > > > > > > > > > lot
>>> > >> > > > > > > > > > > > > >> of
>>> > >> > > > > > > > > > > > > >> > > > >>> hardware
>>> > >> > > > > > > > > > > > > >> > > > >>> > with it.
>>> > >> > > > > > > > > > > > > >> > > > >>> >
>>> > >> > > > > > > > > > > > > >> > > > >>> > Thank you.
>>> > >> > > > > > > > > > > > > >> > > > >>> >
>>> > >> > > > > > > > > > > > > >> > > > >>> > On Sat, Jun 9, 2018 at 14:14
>>> > >> Dongjin
>>> > >> > > Lee <
>>> > >> > > > > > > > > > > > dongjin@apache.org>
>>> > >> > > > > > > > > > > > > >> > > wrote:
>>> > >> > > > > > > > > > > > > >> > > > >>> >
>>> > >> > > > > > > > > > > > > >> > > > >>> > > Since there are no
>>> responses
>>> > for
>>> > >> a
>>> > >> > > > week, I
>>> > >> > > > > > > > decided
>>> > >> > > > > > > > > > to
>>> > >> > > > > > > > > > > > > >> > reinitiate
>>> > >> > > > > > > > > > > > > >> > > > the
>>> > >> > > > > > > > > > > > > >> > > > >>> > > discussion thread.
>>> > >> > > > > > > > > > > > > >> > > > >>> > >
>>> > >> > > > > >
>>
>>

-- 
*Dongjin Lee*

*A hitchhiker in the mathematical world.*

*github:  <http://goog_969573159/>github.com/dongjinleekr
<http://github.com/dongjinleekr>linkedin: kr.linkedin.com/in/dongjinleekr
<http://kr.linkedin.com/in/dongjinleekr>slideshare:
www.slideshare.net/dongjinleekr
<http://www.slideshare.net/dongjinleekr>*

Re: [DISCUSS] KIP-110: Add Codec for ZStandard Compression (Updated)

[Dongjin Lee <do...@apache.org>]

Hi Ismael,

Sure. Thanks.

- Dongjin

On Wed, Sep 12, 2018 at 11:56 PM Ismael Juma <is...@juma.me.uk> wrote:

> Dongjin, can you please start a vote?
>
> Ismael
>
> On Sun, Sep 9, 2018 at 11:15 PM Dongjin Lee <do...@apache.org> wrote:
>
> > Hi Jason,
> >
> > You are right. Explicit statements are always better. I updated the
> > document following your suggestion.
> >
> > @Magnus
> >
> > Thanks for the inspection. It seems like a new error code is not a
> problem.
> >
> > Thanks,
> > Dongjin
> >
> > On Fri, Sep 7, 2018 at 3:23 AM Jason Gustafson <ja...@confluent.io>
> wrote:
> >
> > > Hi Dongjin,
> > >
> > > The KIP looks good to me. I'd suggest starting a vote. A couple minor
> > > points that might be worth calling out explicitly in the compatibility
> > > section:
> > >
> > > 1. Zstd will only be allowed for the bumped produce API. For older
> > > versions, we return UNSUPPORTED_COMPRESSION_TYPE regardless of the
> > message
> > > format.
> > > 2. Down-conversion of zstd-compressed records will not be supported.
> > > Instead we will return UNSUPPORTED_COMPRESSION_TYPE.
> > >
> > > Does that sound right?
> > >
> > > Thanks,
> > > Jason
> > >
> > > On Thu, Sep 6, 2018 at 1:45 AM, Magnus Edenhill <ma...@edenhill.se>
> > > wrote:
> > >
> > > > > Ismael wrote:
> > > > > Jason, that's an interesting point regarding the Java client. Do we
> > > know
> > > > > what clients in other languages do in these cases?
> > > >
> > > > librdkafka (and its bindings) passes unknown/future errors through to
> > the
> > > > application, the error code remains intact while
> > > > the error string will be set to something like "Err-123?", which
> isn't
> > > very
> > > > helpful to the user but it at least
> > > > preserves the original error code for further troubleshooting.
> > > > For the producer any unknown error returned in the ProduceResponse
> will
> > > be
> > > > considered a permanent delivery failure (no retries),
> > > > and for the consumer any unknown FetchResponse errors will propagate
> > > > directly to the application, trigger a fetch backoff, and then
> > > > continue fetching past that offset.
> > > >
> > > > So, from the client's perspective it is not really a problem if new
> > error
> > > > codes are added to older API versions.
> > > >
> > > > /Magnus
> > > >
> > > >
> > > > Den tors 6 sep. 2018 kl 09:45 skrev Dongjin Lee <dongjin@apache.org
> >:
> > > >
> > > > > I updated the KIP page
> > > > > <
> > > > > https://cwiki.apache.org/confluence/display/KAFKA/KIP-
> > > > 110%3A+Add+Codec+for+ZStandard+Compression
> > > > > >
> > > > > following the discussion here. Please take a look when you are
> free.
> > > > > If you have any opinion, don't hesitate to give me a message.
> > > > >
> > > > > Best,
> > > > > Dongjin
> > > > >
> > > > > On Fri, Aug 31, 2018 at 11:35 PM Dongjin Lee <do...@apache.org>
> > > wrote:
> > > > >
> > > > > > I just updated the draft implementation[^1], rebasing against the
> > > > latest
> > > > > > trunk and implementing error routine (i.e., Error code 74 for
> > > > > > UnsupportedCompressionTypeException.) Since we decided to
> disallow
> > > all
> > > > > > fetch request below version 2.1.0 for the topics specifying
> > > ZStandard,
> > > > I
> > > > > > added an error logic only.
> > > > > >
> > > > > > Please have a look when you are free.
> > > > > >
> > > > > > Thanks,
> > > > > > Dongjin
> > > > > >
> > > > > > [^1]: Please check the last commit here:
> > > > > > https://github.com/apache/kafka/pull/2267
> > > > > >
> > > > > > On Thu, Aug 23, 2018, 8:55 AM Dongjin Lee <do...@apache.org>
> > > wrote:
> > > > > >
> > > > > >> Jason,
> > > > > >>
> > > > > >> Great. +1 for UNSUPPORTED_COMPRESSION_TYPE.
> > > > > >>
> > > > > >> Best,
> > > > > >> Dongjin
> > > > > >>
> > > > > >> On Thu, Aug 23, 2018 at 8:19 AM Jason Gustafson <
> > jason@confluent.io
> > > >
> > > > > >> wrote:
> > > > > >>
> > > > > >>> Hey Dongjin,
> > > > > >>>
> > > > > >>> Yeah that's right. For what it's worth, librdkafka also appears
> > to
> > > > > handle
> > > > > >>> unexpected error codes. I expect that most client
> implementations
> > > > would
> > > > > >>> either pass through the raw type or convert to an enum using
> > > > something
> > > > > >>> like
> > > > > >>> what the java client does. Since we're expecting the client to
> > fail
> > > > > >>> anyway,
> > > > > >>> I'm probably in favor of using the UNSUPPORTED_COMPRESSION_TYPE
> > > error
> > > > > >>> code.
> > > > > >>>
> > > > > >>> -Jason
> > > > > >>>
> > > > > >>> On Wed, Aug 22, 2018 at 1:46 AM, Dongjin Lee <
> dongjin@apache.org
> > >
> > > > > wrote:
> > > > > >>>
> > > > > >>> > Jason and Ismael,
> > > > > >>> >
> > > > > >>> > It seems like the only thing we need to regard if we define a
> > new
> > > > > error
> > > > > >>> > code (i.e., UNSUPPORTED_COMPRESSION_TYPE) would be the
> > > > implementation
> > > > > >>> of
> > > > > >>> > the other language clients, right? At least, this strategy
> > causes
> > > > any
> > > > > >>> > problem for Java client. Do I understand correctly?
> > > > > >>> >
> > > > > >>> > Thanks,
> > > > > >>> > Dongjin
> > > > > >>> >
> > > > > >>> > On Wed, Aug 22, 2018 at 5:43 PM Dongjin Lee <
> > dongjin@apache.org>
> > > > > >>> wrote:
> > > > > >>> >
> > > > > >>> > > Jason,
> > > > > >>> > >
> > > > > >>> > > > I think we would only use this error code when we /know/
> > that
> > > > > zstd
> > > > > >>> was
> > > > > >>> > > in use and the client doesn't support it? This is true if
> > > either
> > > > 1)
> > > > > >>> the
> > > > > >>> > > message needs down-conversion and we encounter a zstd
> > > compressed
> > > > > >>> message,
> > > > > >>> > > or 2) if the topic is explicitly configured to use zstd.
> > > > > >>> > >
> > > > > >>> > > Yes, it is right. And you know, the case 1 includes 1.a)
> old
> > > > > clients'
> > > > > >>> > > request v0, v1 records or 1.b) implicit zstd, the
> compression
> > > > type
> > > > > of
> > > > > >>> > > "producer" with Zstd compressed data.
> > > > > >>> > >
> > > > > >>> > > > However, if the compression type is set to "producer,"
> then
> > > the
> > > > > >>> fetched
> > > > > >>> > > data may or may not be compressed with zstd. In this case,
> we
> > > > > return
> > > > > >>> the
> > > > > >>> > > data to the client and expect it to fail parsing. Is that
> > > > correct?
> > > > > >>> > >
> > > > > >>> > > Exactly.
> > > > > >>> > >
> > > > > >>> > > Following your message, I reviewed the implementation of
> > > > > >>> > > `KafkaApis#handleFetchRequest,` which handles the fetch
> > > request.
> > > > > And
> > > > > >>> > found
> > > > > >>> > > that the information we can use is like the following:
> > > > > >>> > >
> > > > > >>> > > 1. Client's fetch request version. (`versionId` variable)
> > > > > >>> > > 2. Log's compression type. (`logConfig` variable)
> > > > > >>> > >
> > > > > >>> > > We can't detect the actual compression type of the data,
> > unless
> > > > we
> > > > > >>> > inspect
> > > > > >>> > > the `RecordBatch` included in the `Records` instance (i.e.,
> > > > > >>> > > `unconvertedRecords` variable.) Since it requires some
> > > > performance
> > > > > >>> issue,
> > > > > >>> > > it is not our option - in short, we can't be sure if given
> > > chunks
> > > > > of
> > > > > >>> data
> > > > > >>> > > are compressed with zstd or not.
> > > > > >>> > >
> > > > > >>> > > So, conclusion: we can return an error in the case of 1.a
> > and 2
> > > > > >>> easily,
> > > > > >>> > > with the information above. In the case 1.b (implicit
> zstd),
> > we
> > > > can
> > > > > >>> just
> > > > > >>> > > return the data by do nothing special and expect it to fail
> > > > > parsing.
> > > > > >>> > >
> > > > > >>> > > Thanks,
> > > > > >>> > > Dongjin
> > > > > >>> > >
> > > > > >>> > > On Wed, Aug 22, 2018 at 12:02 PM Ismael Juma <
> > > ismaelj@gmail.com>
> > > > > >>> wrote:
> > > > > >>> > >
> > > > > >>> > >> Jason, that's an interesting point regarding the Java
> > client.
> > > Do
> > > > > we
> > > > > >>> know
> > > > > >>> > >> what clients in other languages do in these cases?
> > > > > >>> > >>
> > > > > >>> > >> Ismael
> > > > > >>> > >>
> > > > > >>> > >> On Tue, 21 Aug 2018, 17:30 Jason Gustafson, <
> > > jason@confluent.io
> > > > >
> > > > > >>> wrote:
> > > > > >>> > >>
> > > > > >>> > >> > Hi Dongjin,
> > > > > >>> > >> >
> > > > > >>> > >> > One of the complications is that old versions of the API
> > > will
> > > > > not
> > > > > >>> > >> expect a
> > > > > >>> > >> > new error code. However, since we expect this to be a
> > fatal
> > > > > error
> > > > > >>> > anyway
> > > > > >>> > >> > for old clients, it may still be more useful to return
> the
> > > > > correct
> > > > > >>> > error
> > > > > >>> > >> > code. For example, the Kafka clients use the following
> > code
> > > to
> > > > > >>> convert
> > > > > >>> > >> the
> > > > > >>> > >> > error code:
> > > > > >>> > >> >
> > > > > >>> > >> >     public static Errors forCode(short code) {
> > > > > >>> > >> >         Errors error = codeToError.get(code);
> > > > > >>> > >> >         if (error != null) {
> > > > > >>> > >> >             return error;
> > > > > >>> > >> >         } else {
> > > > > >>> > >> >             log.warn("Unexpected error code: {}.",
> code);
> > > > > >>> > >> >             return UNKNOWN_SERVER_ERROR;
> > > > > >>> > >> >         }
> > > > > >>> > >> >     }
> > > > > >>> > >> >
> > > > > >>> > >> > If we return an unsupported error code, it will be
> > converted
> > > > to
> > > > > an
> > > > > >>> > >> UNKNOWN
> > > > > >>> > >> > error, but at least we will get the message in the log
> > with
> > > > the
> > > > > >>> > correct
> > > > > >>> > >> > code. That seems preferable to returning a misleading
> > error
> > > > > code.
> > > > > >>> So I
> > > > > >>> > >> > wonder if we can use the new
> UNSUPPORTED_COMPRESSION_TYPE
> > > > error
> > > > > >>> even
> > > > > >>> > for
> > > > > >>> > >> > older versions.
> > > > > >>> > >> >
> > > > > >>> > >> > Also, one question just to check my understanding. I
> think
> > > we
> > > > > >>> would
> > > > > >>> > only
> > > > > >>> > >> > use this error code when we /know/ that zstd was in use
> > and
> > > > the
> > > > > >>> client
> > > > > >>> > >> > doesn't support it? This is true if either 1) the
> message
> > > > needs
> > > > > >>> > >> > down-conversion and we encounter a zstd compressed
> > message,
> > > or
> > > > > 2)
> > > > > >>> if
> > > > > >>> > the
> > > > > >>> > >> > topic is explicitly configured to use zstd. However, if
> > the
> > > > > >>> > compression
> > > > > >>> > >> > type is set to "producer," then the fetched data may or
> > may
> > > > not
> > > > > be
> > > > > >>> > >> > compressed with zstd. In this case, we return the data
> to
> > > the
> > > > > >>> client
> > > > > >>> > and
> > > > > >>> > >> > expect it to fail parsing. Is that correct?
> > > > > >>> > >> >
> > > > > >>> > >> > Thanks,
> > > > > >>> > >> > Jason
> > > > > >>> > >> >
> > > > > >>> > >> >
> > > > > >>> > >> >
> > > > > >>> > >> > On Tue, Aug 21, 2018 at 9:08 AM, Dongjin Lee <
> > > > > dongjin@apache.org>
> > > > > >>> > >> wrote:
> > > > > >>> > >> >
> > > > > >>> > >> > > Ismael, Jason and all,
> > > > > >>> > >> > >
> > > > > >>> > >> > > I rewrote the backward compatibility strategy & its
> > > > > alternatives
> > > > > >>> > like
> > > > > >>> > >> > > following, based on Ismael & Jason's comments. Since
> it
> > is
> > > > not
> > > > > >>> > >> updated to
> > > > > >>> > >> > > the wiki yet, don't hesitate to give me a message if
> you
> > > > have
> > > > > >>> any
> > > > > >>> > >> opinion
> > > > > >>> > >> > > on it.
> > > > > >>> > >> > >
> > > > > >>> > >> > > ```
> > > > > >>> > >> > > *Backward Compatibility*
> > > > > >>> > >> > >
> > > > > >>> > >> > > We need to establish some backward-compatibility
> > strategy
> > > > for
> > > > > >>> the
> > > > > >>> > >> case an
> > > > > >>> > >> > > old client subscribes a topic using ZStandard
> implicitly
> > > > > (i.e.,
> > > > > >>> > >> > > 'compression.type' configuration of given topic is
> > > > 'producer'
> > > > > >>> and
> > > > > >>> > the
> > > > > >>> > >> > > producer compressed the records with ZStandard). We
> have
> > > the
> > > > > >>> > following
> > > > > >>> > >> > > options for this situation:
> > > > > >>> > >> > >
> > > > > >>> > >> > > *A. Support ZStandard to the old clients which can
> > > > understand
> > > > > >>> v0, v1
> > > > > >>> > >> > > messages only.*
> > > > > >>> > >> > >
> > > > > >>> > >> > > This strategy necessarily requires the down-conversion
> > of
> > > v2
> > > > > >>> message
> > > > > >>> > >> > > compressed with Zstandard into v0 or v1 messages,
> which
> > > > means
> > > > > a
> > > > > >>> > >> > > considerable performance degradation. So we rejected
> > this
> > > > > >>> strategy.
> > > > > >>> > >> > >
> > > > > >>> > >> > > *B. Bump the API version and support only v2-available
> > > > > clients*
> > > > > >>> > >> > >
> > > > > >>> > >> > > With this approach, we can message the old clients
> that
> > > they
> > > > > >>> are old
> > > > > >>> > >> and
> > > > > >>> > >> > > should be upgraded. However, there are still several
> > > options
> > > > > >>> for the
> > > > > >>> > >> > Error
> > > > > >>> > >> > > code.
> > > > > >>> > >> > >
> > > > > >>> > >> > > *B.1. INVALID_REQUEST (42)*
> > > > > >>> > >> > >
> > > > > >>> > >> > > This option gives the client so little information;
> the
> > > user
> > > > > >>> can be
> > > > > >>> > >> > > confused about why the client worked correctly in the
> > past
> > > > > >>> suddenly
> > > > > >>> > >> > > encounters a problem. So we rejected this strategy.
> > > > > >>> > >> > >
> > > > > >>> > >> > > *B.2. CORRUPT_MESSAGE (2)*
> > > > > >>> > >> > >
> > > > > >>> > >> > > This option gives inaccurate information; the user can
> > be
> > > > > >>> surprised
> > > > > >>> > >> and
> > > > > >>> > >> > > misunderstand that the log files are broken in some
> way.
> > > So
> > > > we
> > > > > >>> > >> rejected
> > > > > >>> > >> > > this strategy.
> > > > > >>> > >> > >
> > > > > >>> > >> > > *B.3 UNSUPPORTED_FOR_MESSAGE_FORMAT (43)*
> > > > > >>> > >> > >
> > > > > >>> > >> > > The advantage of this approach is we don't need to
> > define
> > > a
> > > > > new
> > > > > >>> > error
> > > > > >>> > >> > code;
> > > > > >>> > >> > > we can reuse it and that's all.
> > > > > >>> > >> > >
> > > > > >>> > >> > > The disadvantage of this approach is that it is also a
> > > > little
> > > > > >>> bit
> > > > > >>> > >> vague;
> > > > > >>> > >> > > This error code is defined as a work for KIP-98[^1]
> and
> > > now
> > > > > >>> returned
> > > > > >>> > >> in
> > > > > >>> > >> > the
> > > > > >>> > >> > > transaction error.
> > > > > >>> > >> > >
> > > > > >>> > >> > > *B.4. UNSUPPORTED_COMPRESSION_TYPE (new)*
> > > > > >>> > >> > >
> > > > > >>> > >> > > The advantage of this approach is that it is clear and
> > > > > provides
> > > > > >>> an
> > > > > >>> > >> exact
> > > > > >>> > >> > > description. The disadvantage is we need to add a new
> > > error
> > > > > >>> code.
> > > > > >>> > >> > > ```
> > > > > >>> > >> > >
> > > > > >>> > >> > > *It seems like what we need to choose is now so clear:
> > > > > >>> > >> > > UNSUPPORTED_FOR_MESSAGE_FORMAT (B.3) or
> > > > > >>> UNSUPPORTED_COMPRESSION_TYPE
> > > > > >>> > >> > > (B.4).*
> > > > > >>> > >> > > The first one doesn't need a new error message but the
> > > > latter
> > > > > is
> > > > > >>> > more
> > > > > >>> > >> > > explicit. Which one do you prefer? Since all of you
> have
> > > > much
> > > > > >>> more
> > > > > >>> > >> > > experience and knowledge than me, I will follow your
> > > > decision.
> > > > > >>> The
> > > > > >>> > >> wiki
> > > > > >>> > >> > > page will be updated following the decision also.
> > > > > >>> > >> > >
> > > > > >>> > >> > > Best,
> > > > > >>> > >> > > Dongjin
> > > > > >>> > >> > >
> > > > > >>> > >> > > [^1]:
> https://issues.apache.org/jira/browse/KAFKA-4990
> > > > > >>> > >> > >
> > > > > >>> > >> > > On Sun, Aug 19, 2018 at 4:58 AM Ismael Juma <
> > > > > ismaelj@gmail.com>
> > > > > >>> > >> wrote:
> > > > > >>> > >> > >
> > > > > >>> > >> > > > Sounds reasonable to me.
> > > > > >>> > >> > > >
> > > > > >>> > >> > > > Ismael
> > > > > >>> > >> > > >
> > > > > >>> > >> > > > On Sat, 18 Aug 2018, 12:20 Jason Gustafson, <
> > > > > >>> jason@confluent.io>
> > > > > >>> > >> > wrote:
> > > > > >>> > >> > > >
> > > > > >>> > >> > > > > Hey Ismael,
> > > > > >>> > >> > > > >
> > > > > >>> > >> > > > > Your summary looks good to me. I think it might
> also
> > > be
> > > > a
> > > > > >>> good
> > > > > >>> > >> idea
> > > > > >>> > >> > to
> > > > > >>> > >> > > > add
> > > > > >>> > >> > > > > a new UNSUPPORTED_COMPRESSION_TYPE error code to
> go
> > > > along
> > > > > >>> with
> > > > > >>> > the
> > > > > >>> > >> > > > version
> > > > > >>> > >> > > > > bumps. We won't be able to use it for old api
> > versions
> > > > > >>> since the
> > > > > >>> > >> > > clients
> > > > > >>> > >> > > > > will not understand it, but we can use it going
> > > forward
> > > > so
> > > > > >>> that
> > > > > >>> > >> we're
> > > > > >>> > >> > > not
> > > > > >>> > >> > > > > stuck in a similar situation with a new message
> > format
> > > > and
> > > > > >>> a new
> > > > > >>> > >> > codec
> > > > > >>> > >> > > to
> > > > > >>> > >> > > > > support. Another option is to use
> > > > UNSUPPORTED_FOR_MESSAGE_
> > > > > >>> > FORMAT,
> > > > > >>> > >> but
> > > > > >>> > >> > > it
> > > > > >>> > >> > > > is
> > > > > >>> > >> > > > > not as explicit.
> > > > > >>> > >> > > > >
> > > > > >>> > >> > > > > -Jason
> > > > > >>> > >> > > > >
> > > > > >>> > >> > > > > On Fri, Aug 17, 2018 at 5:19 PM, Ismael Juma <
> > > > > >>> ismael@juma.me.uk
> > > > > >>> > >
> > > > > >>> > >> > > wrote:
> > > > > >>> > >> > > > >
> > > > > >>> > >> > > > > > Hi Dongjin and Jason,
> > > > > >>> > >> > > > > >
> > > > > >>> > >> > > > > > I would agree. My summary:
> > > > > >>> > >> > > > > >
> > > > > >>> > >> > > > > > 1. Support zstd with message format 2 only.
> > > > > >>> > >> > > > > > 2. Bump produce and fetch request versions.
> > > > > >>> > >> > > > > > 3. Provide broker errors whenever possible based
> > on
> > > > the
> > > > > >>> > request
> > > > > >>> > >> > > version
> > > > > >>> > >> > > > > and
> > > > > >>> > >> > > > > > rely on clients for the cases where the broker
> > can't
> > > > > >>> validate
> > > > > >>> > >> > > > efficiently
> > > > > >>> > >> > > > > > (example message format 2 consumer that supports
> > the
> > > > > >>> latest
> > > > > >>> > >> fetch
> > > > > >>> > >> > > > version
> > > > > >>> > >> > > > > > but doesn't support zstd).
> > > > > >>> > >> > > > > >
> > > > > >>> > >> > > > > > If there's general agreement on this, I suggest
> we
> > > > > update
> > > > > >>> the
> > > > > >>> > >> KIP
> > > > > >>> > >> > to
> > > > > >>> > >> > > > > state
> > > > > >>> > >> > > > > > the proposal and to move the rejected options to
> > its
> > > > own
> > > > > >>> > >> section.
> > > > > >>> > >> > And
> > > > > >>> > >> > > > > then
> > > > > >>> > >> > > > > > start a vote!
> > > > > >>> > >> > > > > >
> > > > > >>> > >> > > > > > Ismael
> > > > > >>> > >> > > > > >
> > > > > >>> > >> > > > > > On Fri, Aug 17, 2018 at 4:00 PM Jason Gustafson
> <
> > > > > >>> > >> > jason@confluent.io>
> > > > > >>> > >> > > > > > wrote:
> > > > > >>> > >> > > > > >
> > > > > >>> > >> > > > > > > Hi Dongjin,
> > > > > >>> > >> > > > > > >
> > > > > >>> > >> > > > > > > Yes, that's a good summary. For clients which
> > > > support
> > > > > >>> v2,
> > > > > >>> > the
> > > > > >>> > >> > > client
> > > > > >>> > >> > > > > can
> > > > > >>> > >> > > > > > > parse the message format and hopefully raise a
> > > > useful
> > > > > >>> error
> > > > > >>> > >> > message
> > > > > >>> > >> > > > > > > indicating the unsupported compression type.
> For
> > > > older
> > > > > >>> > >> clients,
> > > > > >>> > >> > our
> > > > > >>> > >> > > > > > options
> > > > > >>> > >> > > > > > > are probably (1) to down-convert to the old
> > format
> > > > > >>> using no
> > > > > >>> > >> > > > compression
> > > > > >>> > >> > > > > > > type, or (2) to return an error code. I'm
> > leaning
> > > > > >>> toward the
> > > > > >>> > >> > latter
> > > > > >>> > >> > > > as
> > > > > >>> > >> > > > > > the
> > > > > >>> > >> > > > > > > simpler solution, but the challenge is
> finding a
> > > > good
> > > > > >>> error
> > > > > >>> > >> code.
> > > > > >>> > >> > > Two
> > > > > >>> > >> > > > > > > possibilities might be INVALID_REQUEST or
> > > > > >>> CORRUPT_MESSAGE.
> > > > > >>> > The
> > > > > >>> > >> > > > downside
> > > > > >>> > >> > > > > > is
> > > > > >>> > >> > > > > > > that old clients probably won't get a helpful
> > > > message.
> > > > > >>> > >> However,
> > > > > >>> > >> > at
> > > > > >>> > >> > > > > least
> > > > > >>> > >> > > > > > > the behavior will be consistent in the sense
> > that
> > > > all
> > > > > >>> > clients
> > > > > >>> > >> > will
> > > > > >>> > >> > > > fail
> > > > > >>> > >> > > > > > if
> > > > > >>> > >> > > > > > > they do not support zstandard.
> > > > > >>> > >> > > > > > >
> > > > > >>> > >> > > > > > > What do you think?
> > > > > >>> > >> > > > > > >
> > > > > >>> > >> > > > > > > Thanks,
> > > > > >>> > >> > > > > > > Jason
> > > > > >>> > >> > > > > > >
> > > > > >>> > >> > > > > > > On Fri, Aug 17, 2018 at 8:08 AM, Dongjin Lee <
> > > > > >>> > >> dongjin@apache.org
> > > > > >>> > >> > >
> > > > > >>> > >> > > > > wrote:
> > > > > >>> > >> > > > > > >
> > > > > >>> > >> > > > > > > > Thanks Jason, I reviewed the down-converting
> > > logic
> > > > > >>> > following
> > > > > >>> > >> > your
> > > > > >>> > >> > > > > > > > explanation.[^1] You mean the following
> > > routines,
> > > > > >>> right?
> > > > > >>> > >> > > > > > > >
> > > > > >>> > >> > > > > > > > -
> > > > > >>> > >> > > > > > > >
> > > > > https://github.com/apache/kafka/blob/trunk/core/src/
> > > > > >>> > >> > > > > > > > main/scala/kafka/server/KafkaApis.scala#L534
> > > > > >>> > >> > > > > > > > -
> > > > > >>> > >> > > > > > > >
> > > > > >>> https://github.com/apache/kafka/blob/trunk/clients/src/
> > > > > >>> > >> > > > > > > > main/java/org/apache/kafka/common/record/
> > > > > >>> > >> > > LazyDownConversionRecords.
> > > > > >>> > >> > > > > > > > java#L165
> > > > > >>> > >> > > > > > > > -
> > > > > >>> > >> > > > > > > >
> > > > > >>> https://github.com/apache/kafka/blob/trunk/clients/src/
> > > > > >>> > >> > > > > > > >
> > > > > >>> > >>
> > main/java/org/apache/kafka/common/record/RecordsUtil.java#L40
> > > > > >>> > >> > > > > > > >
> > > > > >>> > >> > > > > > > > It seems like your stance is like following:
> > > > > >>> > >> > > > > > > >
> > > > > >>> > >> > > > > > > > 1. In principle, Kafka does not change the
> > > > > compression
> > > > > >>> > codec
> > > > > >>> > >> > when
> > > > > >>> > >> > > > > > > > down-converting, since it requires
> inspecting
> > > the
> > > > > >>> fetched
> > > > > >>> > >> data,
> > > > > >>> > >> > > > which
> > > > > >>> > >> > > > > > is
> > > > > >>> > >> > > > > > > > expensive.
> > > > > >>> > >> > > > > > > > 2. However, there are some cases the fetched
> > > data
> > > > is
> > > > > >>> > >> inspected
> > > > > >>> > >> > > > > anyway.
> > > > > >>> > >> > > > > > In
> > > > > >>> > >> > > > > > > > this case, we can provide compression
> > conversion
> > > > > from
> > > > > >>> > >> Zstandard
> > > > > >>> > >> > > to
> > > > > >>> > >> > > > > > > > classical ones[^2].
> > > > > >>> > >> > > > > > > >
> > > > > >>> > >> > > > > > > > And from what I understand, the cases where
> > the
> > > > > client
> > > > > >>> > >> without
> > > > > >>> > >> > > > > > ZStandard
> > > > > >>> > >> > > > > > > > support receives ZStandard compressed
> records
> > > can
> > > > be
> > > > > >>> > >> organized
> > > > > >>> > >> > > into
> > > > > >>> > >> > > > > two
> > > > > >>> > >> > > > > > > > cases:
> > > > > >>> > >> > > > > > > >
> > > > > >>> > >> > > > > > > > a. The 'compression.type' configuration of
> > given
> > > > > >>> topic is
> > > > > >>> > >> > > > 'producer'
> > > > > >>> > >> > > > > > and
> > > > > >>> > >> > > > > > > > the producer compressed the records with
> > > > ZStandard.
> > > > > >>> (that
> > > > > >>> > >> is,
> > > > > >>> > >> > > using
> > > > > >>> > >> > > > > > > > ZStandard implicitly.)
> > > > > >>> > >> > > > > > > > b.  The 'compression.type' configuration of
> > > given
> > > > > >>> topic is
> > > > > >>> > >> > > 'zstd';
> > > > > >>> > >> > > > > that
> > > > > >>> > >> > > > > > > is,
> > > > > >>> > >> > > > > > > > using ZStandard explicitly.
> > > > > >>> > >> > > > > > > >
> > > > > >>> > >> > > > > > > > As you stated, we don't have to handle the
> > case
> > > b
> > > > > >>> > specially.
> > > > > >>> > >> > So,
> > > > > >>> > >> > > It
> > > > > >>> > >> > > > > > seems
> > > > > >>> > >> > > > > > > > like we can narrow the focus of the problem
> by
> > > > > joining
> > > > > >>> > case
> > > > > >>> > >> 1
> > > > > >>> > >> > and
> > > > > >>> > >> > > > > case
> > > > > >>> > >> > > > > > b
> > > > > >>> > >> > > > > > > > like the following:
> > > > > >>> > >> > > > > > > >
> > > > > >>> > >> > > > > > > > > Given the topic with 'producer' as its
> > > > > >>> > 'compression.type'
> > > > > >>> > >> > > > > > > configuration,
> > > > > >>> > >> > > > > > > > ZStandard compressed records and old client
> > > > without
> > > > > >>> > >> ZStandard,
> > > > > >>> > >> > is
> > > > > >>> > >> > > > > there
> > > > > >>> > >> > > > > > > any
> > > > > >>> > >> > > > > > > > case we need to inspect the records and can
> > > change
> > > > > the
> > > > > >>> > >> > > compression
> > > > > >>> > >> > > > > > type?
> > > > > >>> > >> > > > > > > If
> > > > > >>> > >> > > > > > > > so, can we provide compression type
> > converting?
> > > > > >>> > >> > > > > > > >
> > > > > >>> > >> > > > > > > > Do I understand correctly?
> > > > > >>> > >> > > > > > > >
> > > > > >>> > >> > > > > > > > Best,
> > > > > >>> > >> > > > > > > > Dongjin
> > > > > >>> > >> > > > > > > >
> > > > > >>> > >> > > > > > > > [^1]: I'm sorry, I found that I was a little
> > bit
> > > > > >>> > >> > misunderstanding
> > > > > >>> > >> > > > how
> > > > > >>> > >> > > > > > API
> > > > > >>> > >> > > > > > > > version works, after reviewing the
> downconvert
> > > > logic
> > > > > >>> & the
> > > > > >>> > >> > > protocol
> > > > > >>> > >> > > > > > > > documentation <
> > > https://kafka.apache.org/protocol
> > > > >.
> > > > > >>> > >> > > > > > > > [^2]: None, Gzip, Snappy, Lz4.
> > > > > >>> > >> > > > > > > >
> > > > > >>> > >> > > > > > > > On Tue, Aug 14, 2018 at 2:16 AM Jason
> > Gustafson
> > > <
> > > > > >>> > >> > > > jason@confluent.io>
> > > > > >>> > >> > > > > > > > wrote:
> > > > > >>> > >> > > > > > > >
> > > > > >>> > >> > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > But in my opinion, since the client will
> > > fail
> > > > > >>> with the
> > > > > >>> > >> API
> > > > > >>> > >> > > > > version,
> > > > > >>> > >> > > > > > > so
> > > > > >>> > >> > > > > > > > we
> > > > > >>> > >> > > > > > > > > > don't need to down-convert the messages
> > > > anyway.
> > > > > >>> Isn't
> > > > > >>> > >> it?
> > > > > >>> > >> > > So, I
> > > > > >>> > >> > > > > > think
> > > > > >>> > >> > > > > > > > we
> > > > > >>> > >> > > > > > > > > > don't care about this case. (I'm sorry,
> I
> > am
> > > > not
> > > > > >>> > >> familiar
> > > > > >>> > >> > > with
> > > > > >>> > >> > > > > > > > > down-convert
> > > > > >>> > >> > > > > > > > > > logic.)
> > > > > >>> > >> > > > > > > > >
> > > > > >>> > >> > > > > > > > >
> > > > > >>> > >> > > > > > > > > Currently the broker down-converts
> > > automatically
> > > > > >>> when it
> > > > > >>> > >> > > receives
> > > > > >>> > >> > > > > an
> > > > > >>> > >> > > > > > > old
> > > > > >>> > >> > > > > > > > > version of the fetch request (a version
> > which
> > > is
> > > > > >>> known
> > > > > >>> > to
> > > > > >>> > >> > > predate
> > > > > >>> > >> > > > > the
> > > > > >>> > >> > > > > > > > > message format in use). Typically when
> > > > > >>> down-converting
> > > > > >>> > the
> > > > > >>> > >> > > > message
> > > > > >>> > >> > > > > > > > format,
> > > > > >>> > >> > > > > > > > > we use the same compression type, but
> there
> > is
> > > > not
> > > > > >>> much
> > > > > >>> > >> point
> > > > > >>> > >> > > in
> > > > > >>> > >> > > > > > doing
> > > > > >>> > >> > > > > > > so
> > > > > >>> > >> > > > > > > > > when we know the client doesn't support
> it.
> > So
> > > > if
> > > > > >>> > >> zstandard
> > > > > >>> > >> > is
> > > > > >>> > >> > > in
> > > > > >>> > >> > > > > > use,
> > > > > >>> > >> > > > > > > > and
> > > > > >>> > >> > > > > > > > > we have to down-convert anyway, then we
> can
> > > > choose
> > > > > >>> to
> > > > > >>> > use
> > > > > >>> > >> a
> > > > > >>> > >> > > > > different
> > > > > >>> > >> > > > > > > > > compression type or no compression type.
> > > > > >>> > >> > > > > > > > >
> > > > > >>> > >> > > > > > > > > From my perspective, there is no
> significant
> > > > > >>> downside to
> > > > > >>> > >> > > bumping
> > > > > >>> > >> > > > > the
> > > > > >>> > >> > > > > > > > > protocol version and it has several
> > potential
> > > > > >>> benefits.
> > > > > >>> > >> > Version
> > > > > >>> > >> > > > > bumps
> > > > > >>> > >> > > > > > > are
> > > > > >>> > >> > > > > > > > > cheap. The main question mark in my mind
> is
> > > > about
> > > > > >>> > >> > > > down-conversion.
> > > > > >>> > >> > > > > > > > Figuring
> > > > > >>> > >> > > > > > > > > out whether down-conversion is needed is
> > hard
> > > > > >>> generally
> > > > > >>> > >> > without
> > > > > >>> > >> > > > > > > > inspecting
> > > > > >>> > >> > > > > > > > > the fetched data, which is expensive. I
> > think
> > > we
> > > > > >>> agree
> > > > > >>> > in
> > > > > >>> > >> > > > principle
> > > > > >>> > >> > > > > > > that
> > > > > >>> > >> > > > > > > > we
> > > > > >>> > >> > > > > > > > > do not want to have to pay this cost
> > generally
> > > > and
> > > > > >>> > prefer
> > > > > >>> > >> the
> > > > > >>> > >> > > > > clients
> > > > > >>> > >> > > > > > > to
> > > > > >>> > >> > > > > > > > > fail when they see an unhandled
> compression
> > > > type.
> > > > > >>> The
> > > > > >>> > >> point I
> > > > > >>> > >> > > was
> > > > > >>> > >> > > > > > > making
> > > > > >>> > >> > > > > > > > is
> > > > > >>> > >> > > > > > > > > that there are some cases where we are
> > either
> > > > > >>> inspecting
> > > > > >>> > >> the
> > > > > >>> > >> > > data
> > > > > >>> > >> > > > > > > anyway
> > > > > >>> > >> > > > > > > > > (because we have to down-convert the
> message
> > > > > >>> format), or
> > > > > >>> > >> we
> > > > > >>> > >> > > have
> > > > > >>> > >> > > > an
> > > > > >>> > >> > > > > > > easy
> > > > > >>> > >> > > > > > > > > way to tell whether zstandard is in use
> (the
> > > > topic
> > > > > >>> has
> > > > > >>> > it
> > > > > >>> > >> > > > > configured
> > > > > >>> > >> > > > > > > > > explicitly). In the latter case, we don't
> > have
> > > > to
> > > > > >>> handle
> > > > > >>> > >> it
> > > > > >>> > >> > > > > > specially.
> > > > > >>> > >> > > > > > > > But
> > > > > >>> > >> > > > > > > > > we do have to decide how we will handle
> > > > > >>> down-conversion
> > > > > >>> > to
> > > > > >>> > >> > > older
> > > > > >>> > >> > > > > > > formats.
> > > > > >>> > >> > > > > > > > >
> > > > > >>> > >> > > > > > > > > -Jason
> > > > > >>> > >> > > > > > > > >
> > > > > >>> > >> > > > > > > > > On Sun, Aug 12, 2018 at 5:15 PM, Dongjin
> > Lee <
> > > > > >>> > >> > > dongjin@apache.org
> > > > > >>> > >> > > > >
> > > > > >>> > >> > > > > > > wrote:
> > > > > >>> > >> > > > > > > > >
> > > > > >>> > >> > > > > > > > > > Colin and Jason,
> > > > > >>> > >> > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > Thanks for your opinions. In
> summarizing,
> > > the
> > > > > >>> Pros and
> > > > > >>> > >> Cons
> > > > > >>> > >> > > of
> > > > > >>> > >> > > > > > > bumping
> > > > > >>> > >> > > > > > > > > > fetch API version are:
> > > > > >>> > >> > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > Cons:
> > > > > >>> > >> > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > - The Broker can't know whether a given
> > > > message
> > > > > >>> batch
> > > > > >>> > is
> > > > > >>> > >> > > > > compressed
> > > > > >>> > >> > > > > > > > with
> > > > > >>> > >> > > > > > > > > > zstd or not.
> > > > > >>> > >> > > > > > > > > > - Need some additional logic for the
> topic
> > > > > >>> explicitly
> > > > > >>> > >> > > > configured
> > > > > >>> > >> > > > > to
> > > > > >>> > >> > > > > > > use
> > > > > >>> > >> > > > > > > > > > zstd.
> > > > > >>> > >> > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > Pros:
> > > > > >>> > >> > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > - The broker doesn't need to conduct
> > > expensive
> > > > > >>> > >> > > down-conversion.
> > > > > >>> > >> > > > > > > > > > - Can message the users to update their
> > > > client.
> > > > > >>> > >> > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > So, opinions for the
> > backward-compatibility
> > > > > >>> policy by
> > > > > >>> > >> far:
> > > > > >>> > >> > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > - A: bump the API version - +2 (Colin,
> > > Jason)
> > > > > >>> > >> > > > > > > > > > - B: leave unchanged - +1 (Viktor)
> > > > > >>> > >> > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > Here are my additional comments:
> > > > > >>> > >> > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > @Colin
> > > > > >>> > >> > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > I greatly appreciate your response. In
> the
> > > > case
> > > > > >>> of the
> > > > > >>> > >> > > > dictionary
> > > > > >>> > >> > > > > > > > > support,
> > > > > >>> > >> > > > > > > > > > of course, this issue should be
> addressed
> > > > later
> > > > > >>> so we
> > > > > >>> > >> don't
> > > > > >>> > >> > > > need
> > > > > >>> > >> > > > > it
> > > > > >>> > >> > > > > > > in
> > > > > >>> > >> > > > > > > > > the
> > > > > >>> > >> > > > > > > > > > first version. You are right - it is not
> > > late
> > > > to
> > > > > >>> try
> > > > > >>> > it
> > > > > >>> > >> > after
> > > > > >>> > >> > > > > some
> > > > > >>> > >> > > > > > > > > > benchmarks. What I mean is, we should
> keep
> > > in
> > > > > >>> mind on
> > > > > >>> > >> that
> > > > > >>> > >> > > > > > potential
> > > > > >>> > >> > > > > > > > > > feature.
> > > > > >>> > >> > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > @Jason
> > > > > >>> > >> > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > You wrote,
> > > > > >>> > >> > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > > Similarly, if we have to down-convert
> > > anyway
> > > > > >>> because
> > > > > >>> > >> the
> > > > > >>> > >> > > > client
> > > > > >>> > >> > > > > > > does
> > > > > >>> > >> > > > > > > > > not
> > > > > >>> > >> > > > > > > > > > understand the message format, then we
> > could
> > > > > also
> > > > > >>> use
> > > > > >>> > a
> > > > > >>> > >> > > > different
> > > > > >>> > >> > > > > > > > > > compression type.
> > > > > >>> > >> > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > But in my opinion, since the client will
> > > fail
> > > > > >>> with the
> > > > > >>> > >> API
> > > > > >>> > >> > > > > version,
> > > > > >>> > >> > > > > > > so
> > > > > >>> > >> > > > > > > > we
> > > > > >>> > >> > > > > > > > > > don't need to down-convert the messages
> > > > anyway.
> > > > > >>> Isn't
> > > > > >>> > >> it?
> > > > > >>> > >> > > So, I
> > > > > >>> > >> > > > > > think
> > > > > >>> > >> > > > > > > > we
> > > > > >>> > >> > > > > > > > > > don't care about this case. (I'm sorry,
> I
> > am
> > > > not
> > > > > >>> > >> familiar
> > > > > >>> > >> > > with
> > > > > >>> > >> > > > > > > > > down-convert
> > > > > >>> > >> > > > > > > > > > logic.)
> > > > > >>> > >> > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > Please give more opinions. Thanks!
> > > > > >>> > >> > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > - Dongjin
> > > > > >>> > >> > > > > > > > > >
> > > > > >>> > >> > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > On Wed, Aug 8, 2018 at 6:41 AM Jason
> > > > Gustafson <
> > > > > >>> > >> > > > > jason@confluent.io
> > > > > >>> > >> > > > > > >
> > > > > >>> > >> > > > > > > > > wrote:
> > > > > >>> > >> > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > > Hey Colin,
> > > > > >>> > >> > > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > > The problem for the fetch API is that
> > the
> > > > > broker
> > > > > >>> > does
> > > > > >>> > >> not
> > > > > >>> > >> > > > > > generally
> > > > > >>> > >> > > > > > > > > know
> > > > > >>> > >> > > > > > > > > > if
> > > > > >>> > >> > > > > > > > > > > a batch was compressed with zstd
> unless
> > it
> > > > > >>> parses
> > > > > >>> > it.
> > > > > >>> > >> I
> > > > > >>> > >> > > think
> > > > > >>> > >> > > > > the
> > > > > >>> > >> > > > > > > > goal
> > > > > >>> > >> > > > > > > > > > here
> > > > > >>> > >> > > > > > > > > > > is to avoid the expensive
> > down-conversion
> > > > that
> > > > > >>> is
> > > > > >>> > >> needed
> > > > > >>> > >> > to
> > > > > >>> > >> > > > > > ensure
> > > > > >>> > >> > > > > > > > > > > compatibility because it is only
> > necessary
> > > > if
> > > > > >>> zstd
> > > > > >>> > is
> > > > > >>> > >> > > > actually
> > > > > >>> > >> > > > > in
> > > > > >>> > >> > > > > > > > use.
> > > > > >>> > >> > > > > > > > > > But
> > > > > >>> > >> > > > > > > > > > > as long as old clients can parse the
> > > message
> > > > > >>> format,
> > > > > >>> > >> they
> > > > > >>> > >> > > > > should
> > > > > >>> > >> > > > > > > get
> > > > > >>> > >> > > > > > > > a
> > > > > >>> > >> > > > > > > > > > > reasonable error if they see an
> > > unsupported
> > > > > >>> > >> compression
> > > > > >>> > >> > > type
> > > > > >>> > >> > > > in
> > > > > >>> > >> > > > > > the
> > > > > >>> > >> > > > > > > > > > > attributes. Basically the onus is on
> > users
> > > > to
> > > > > >>> ensure
> > > > > >>> > >> that
> > > > > >>> > >> > > > their
> > > > > >>> > >> > > > > > > > > consumers
> > > > > >>> > >> > > > > > > > > > > have been updated prior to using zstd.
> > It
> > > > > seems
> > > > > >>> > like a
> > > > > >>> > >> > > > > reasonable
> > > > > >>> > >> > > > > > > > > > tradeoff
> > > > > >>> > >> > > > > > > > > > > to me. There are a couple cases that
> > might
> > > > be
> > > > > >>> worth
> > > > > >>> > >> > > thinking
> > > > > >>> > >> > > > > > > through:
> > > > > >>> > >> > > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > > 1. If a topic is explicitly configured
> > to
> > > > use
> > > > > >>> zstd,
> > > > > >>> > >> then
> > > > > >>> > >> > we
> > > > > >>> > >> > > > > don't
> > > > > >>> > >> > > > > > > > need
> > > > > >>> > >> > > > > > > > > to
> > > > > >>> > >> > > > > > > > > > > check the fetched data for the
> > compression
> > > > > type
> > > > > >>> to
> > > > > >>> > >> know
> > > > > >>> > >> > if
> > > > > >>> > >> > > we
> > > > > >>> > >> > > > > > need
> > > > > >>> > >> > > > > > > > > > > down-conversion. If we did bump the
> > Fetch
> > > > API
> > > > > >>> > version,
> > > > > >>> > >> > then
> > > > > >>> > >> > > > we
> > > > > >>> > >> > > > > > > could
> > > > > >>> > >> > > > > > > > > > handle
> > > > > >>> > >> > > > > > > > > > > this case by either down-converting
> > using
> > > a
> > > > > >>> > different
> > > > > >>> > >> > > > > compression
> > > > > >>> > >> > > > > > > > type
> > > > > >>> > >> > > > > > > > > or
> > > > > >>> > >> > > > > > > > > > > returning an error.
> > > > > >>> > >> > > > > > > > > > > 2. Similarly, if we have to
> down-convert
> > > > > anyway
> > > > > >>> > >> because
> > > > > >>> > >> > the
> > > > > >>> > >> > > > > > client
> > > > > >>> > >> > > > > > > > does
> > > > > >>> > >> > > > > > > > > > not
> > > > > >>> > >> > > > > > > > > > > understand the message format, then we
> > > could
> > > > > >>> also
> > > > > >>> > use
> > > > > >>> > >> a
> > > > > >>> > >> > > > > different
> > > > > >>> > >> > > > > > > > > > > compression type.
> > > > > >>> > >> > > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > > For the produce API, I think it's
> > > reasonable
> > > > > to
> > > > > >>> bump
> > > > > >>> > >> the
> > > > > >>> > >> > > api
> > > > > >>> > >> > > > > > > version.
> > > > > >>> > >> > > > > > > > > > This
> > > > > >>> > >> > > > > > > > > > > can be used by clients to check
> whether
> > a
> > > > > broker
> > > > > >>> > >> supports
> > > > > >>> > >> > > > zstd.
> > > > > >>> > >> > > > > > For
> > > > > >>> > >> > > > > > > > > > > example, we might support a list of
> > > > preferred
> > > > > >>> > >> compression
> > > > > >>> > >> > > > types
> > > > > >>> > >> > > > > > in
> > > > > >>> > >> > > > > > > > the
> > > > > >>> > >> > > > > > > > > > > producer and we could use the broker
> to
> > > > detect
> > > > > >>> which
> > > > > >>> > >> > > version
> > > > > >>> > >> > > > to
> > > > > >>> > >> > > > > > > use.
> > > > > >>> > >> > > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > > -Jason
> > > > > >>> > >> > > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > > On Tue, Aug 7, 2018 at 1:32 PM, Colin
> > > > McCabe <
> > > > > >>> > >> > > > > cmccabe@apache.org
> > > > > >>> > >> > > > > > >
> > > > > >>> > >> > > > > > > > > wrote:
> > > > > >>> > >> > > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > > > Thanks for bumping this, Dongjin.
> > ZStd
> > > > is a
> > > > > >>> good
> > > > > >>> > >> > > > compression
> > > > > >>> > >> > > > > > > codec
> > > > > >>> > >> > > > > > > > > > and I
> > > > > >>> > >> > > > > > > > > > > > hope we can get this support in
> soon!
> > > > > >>> > >> > > > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > > > I would say we can just bump the API
> > > > version
> > > > > >>> to
> > > > > >>> > >> > indicate
> > > > > >>> > >> > > > that
> > > > > >>> > >> > > > > > > ZStd
> > > > > >>> > >> > > > > > > > > > > support
> > > > > >>> > >> > > > > > > > > > > > is expected in new clients.  We
> > probably
> > > > > need
> > > > > >>> some
> > > > > >>> > >> way
> > > > > >>> > >> > of
> > > > > >>> > >> > > > > > > > indicating
> > > > > >>> > >> > > > > > > > > to
> > > > > >>> > >> > > > > > > > > > > the
> > > > > >>> > >> > > > > > > > > > > > older clients that they can't
> consume
> > > the
> > > > > >>> > >> partitions,
> > > > > >>> > >> > as
> > > > > >>> > >> > > > > well.
> > > > > >>> > >> > > > > > > > > Perhaps
> > > > > >>> > >> > > > > > > > > > > we
> > > > > >>> > >> > > > > > > > > > > > can use the
> > > UNSUPPORTED_FOR_MESSAGE_FORMAT
> > > > > >>> error?
> > > > > >>> > >> > > > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > > > The license thing seems
> > straightforward
> > > --
> > > > > >>> it's
> > > > > >>> > >> just a
> > > > > >>> > >> > > > matter
> > > > > >>> > >> > > > > > of
> > > > > >>> > >> > > > > > > > > adding
> > > > > >>> > >> > > > > > > > > > > > the text to the right files as per
> ASF
> > > > > >>> guidelines.
> > > > > >>> > >> > > > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > > > With regard to the dictionary
> support,
> > > do
> > > > we
> > > > > >>> > really
> > > > > >>> > >> > need
> > > > > >>> > >> > > > that
> > > > > >>> > >> > > > > > in
> > > > > >>> > >> > > > > > > > the
> > > > > >>> > >> > > > > > > > > > > first
> > > > > >>> > >> > > > > > > > > > > > version?  Hopefully message batches
> > are
> > > > big
> > > > > >>> enough
> > > > > >>> > >> that
> > > > > >>> > >> > > > this
> > > > > >>> > >> > > > > > > isn't
> > > > > >>> > >> > > > > > > > > > > needed.
> > > > > >>> > >> > > > > > > > > > > > Some benchmarks might help here.
> > > > > >>> > >> > > > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > > > best,
> > > > > >>> > >> > > > > > > > > > > > Colin
> > > > > >>> > >> > > > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > > > On Tue, Aug 7, 2018, at 08:02,
> Dongjin
> > > Lee
> > > > > >>> wrote:
> > > > > >>> > >> > > > > > > > > > > > > As Kafka 2.0.0 was released, let's
> > > > reboot
> > > > > >>> this
> > > > > >>> > >> issue,
> > > > > >>> > >> > > > > KIP-110
> > > > > >>> > >> > > > > > > > > > > > > <https://cwiki.apache.org/
> > > > > >>> > >> > > confluence/display/KAFKA/KIP-
> > > > > >>> > >> > > > > > > > > > > > 110%3A+Add+Codec+for+
> > > > ZStandard+Compression>
> > > > > >>> > >> > > > > > > > > > > > > .
> > > > > >>> > >> > > > > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > > > > For newcomers, Here is some
> summary
> > of
> > > > the
> > > > > >>> > >> history:
> > > > > >>> > >> > > > KIP-110
> > > > > >>> > >> > > > > > was
> > > > > >>> > >> > > > > > > > > > > > originally
> > > > > >>> > >> > > > > > > > > > > > > worked for the issue KAFKA-4514
> but,
> > > it
> > > > > >>> lacked
> > > > > >>> > >> > > benchmark
> > > > > >>> > >> > > > > > > results
> > > > > >>> > >> > > > > > > > to
> > > > > >>> > >> > > > > > > > > > get
> > > > > >>> > >> > > > > > > > > > > > the
> > > > > >>> > >> > > > > > > > > > > > > agreement of the community. Later,
> > > Ivan
> > > > > >>> Babrou
> > > > > >>> > and
> > > > > >>> > >> > some
> > > > > >>> > >> > > > > other
> > > > > >>> > >> > > > > > > > users
> > > > > >>> > >> > > > > > > > > > who
> > > > > >>> > >> > > > > > > > > > > > > adopted the patch provided their
> > > > excellent
> > > > > >>> > >> > performance
> > > > > >>> > >> > > > > report
> > > > > >>> > >> > > > > > > > which
> > > > > >>> > >> > > > > > > > > > is
> > > > > >>> > >> > > > > > > > > > > > now
> > > > > >>> > >> > > > > > > > > > > > > included in the KIP, but it
> > postponed
> > > > > again
> > > > > >>> > >> because
> > > > > >>> > >> > of
> > > > > >>> > >> > > > the
> > > > > >>> > >> > > > > > > > > community
> > > > > >>> > >> > > > > > > > > > > was
> > > > > >>> > >> > > > > > > > > > > > > busy for 2.0.0 release. It is why
> I
> > > now
> > > > > >>> reboot
> > > > > >>> > >> this
> > > > > >>> > >> > > > issue.
> > > > > >>> > >> > > > > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > > > > The following is the current
> status
> > of
> > > > the
> > > > > >>> > >> feature:
> > > > > >>> > >> > You
> > > > > >>> > >> > > > can
> > > > > >>> > >> > > > > > > check
> > > > > >>> > >> > > > > > > > > the
> > > > > >>> > >> > > > > > > > > > > > > current draft implementation here
> > > > > >>> > >> > > > > > > > > > > > > <
> > > > > https://github.com/apache/kafka/pull/2267>.
> > > > > >>> It
> > > > > >>> > >> is
> > > > > >>> > >> > > based
> > > > > >>> > >> > > > > on
> > > > > >>> > >> > > > > > > zstd
> > > > > >>> > >> > > > > > > > > > 1.3.5
> > > > > >>> > >> > > > > > > > > > > > and
> > > > > >>> > >> > > > > > > > > > > > > periodically rebased onto the
> latest
> > > > > >>> trunk[^1].
> > > > > >>> > >> > > > > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > > > > The issues that should be
> addressed
> > is
> > > > > like
> > > > > >>> > >> > following:
> > > > > >>> > >> > > > > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > > > > *1. Backward Compatibility*
> > > > > >>> > >> > > > > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > > > > To support old consumers, we need
> to
> > > > take
> > > > > a
> > > > > >>> > >> strategy
> > > > > >>> > >> > to
> > > > > >>> > >> > > > > > handle
> > > > > >>> > >> > > > > > > > the
> > > > > >>> > >> > > > > > > > > > old
> > > > > >>> > >> > > > > > > > > > > > > consumers. Current candidates are:
> > > > > >>> > >> > > > > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > > > > - Bump API version
> > > > > >>> > >> > > > > > > > > > > > > - Leave unchanged: let the old
> > clients
> > > > > fail.
> > > > > >>> > >> > > > > > > > > > > > > - Improve the error messages:
> > > > > >>> > >> > > > > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > > > > *2. Dictionary Support*
> > > > > >>> > >> > > > > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > > > > To support zstd's dictionary
> feature
> > > in
> > > > > the
> > > > > >>> > future
> > > > > >>> > >> > (if
> > > > > >>> > >> > > > > > needed),
> > > > > >>> > >> > > > > > > > we
> > > > > >>> > >> > > > > > > > > > need
> > > > > >>> > >> > > > > > > > > > > > to
> > > > > >>> > >> > > > > > > > > > > > > sketch how it should be and leave
> > some
> > > > > room
> > > > > >>> for
> > > > > >>> > >> it.
> > > > > >>> > >> > As
> > > > > >>> > >> > > of
> > > > > >>> > >> > > > > > now,
> > > > > >>> > >> > > > > > > > > there
> > > > > >>> > >> > > > > > > > > > > has
> > > > > >>> > >> > > > > > > > > > > > > been no discussion on this topic
> > yet.
> > > > > >>> > >> > > > > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > > > > *3. License*
> > > > > >>> > >> > > > > > > > > > > > >
> > > > > >>> > >> > > > > > > > > > > > > To use this feature, we need to
> add
> > > > > license
> > > > > >>> of
> > > > > >>> > >> zstd
> > > > > >>> > >> > and
> > > > > >>> > >> > > > > > > zstd-jni
> > > > > >>> > >> > > > > > > > to
> > > > > >>> > >> > > > > > > > > > the
> > > > > >>> > >> > > > > > > > > > > > > project. (Thanks to Viktor Somogyi
> > for
> > > > > >>> raising
> > > > > >>> > >> this
> > > > > >>> > >> > > > issue!)
> > > > > >>> > >> > > > > > It
> > > > > >>> > >> > > > > > > > > seems
> > > > > >>> > >> > > > > > > > > > > like
> > > > > >>> > >> > > > > > > > > > > > > what Apache Spark did would be a
> > good
> > > > > >>> example
> > > > > >>> > but
> > > > > >>> > >> > there
> > > > > >>> > >> > > > has
> > > > > >>> > >> > > > > > > been
> > > > > >>> > >> > > > > > > > no
> > > > > >>> > >> > > > > > > > > > > > > discussion yet.
> > > > > >>> > >> > > > > > > > > > > > >
> > > > > >>> > >> > > > >



-- 
*Dongjin Lee*

*A hitchhiker in the mathematical world.*

*github:  <http://goog_969573159/>github.com/dongjinleekr
<http://github.com/dongjinleekr>linkedin: kr.linkedin.com/in/dongjinleekr
<http://kr.linkedin.com/in/dongjinleekr>slideshare:
www.slideshare.net/dongjinleekr
<http://www.slideshare.net/dongjinleekr>*

Re: [DISCUSS] KIP-110: Add Codec for ZStandard Compression (Updated)

[Ismael Juma <is...@juma.me.uk>]

Dongjin, can you please start a vote?

Ismael

On Sun, Sep 9, 2018 at 11:15 PM Dongjin Lee <do...@apache.org> wrote:

> Hi Jason,
>
> You are right. Explicit statements are always better. I updated the
> document following your suggestion.
>
> @Magnus
>
> Thanks for the inspection. It seems like a new error code is not a problem.
>
> Thanks,
> Dongjin
>
> On Fri, Sep 7, 2018 at 3:23 AM Jason Gustafson <ja...@confluent.io> wrote:
>
> > Hi Dongjin,
> >
> > The KIP looks good to me. I'd suggest starting a vote. A couple minor
> > points that might be worth calling out explicitly in the compatibility
> > section:
> >
> > 1. Zstd will only be allowed for the bumped produce API. For older
> > versions, we return UNSUPPORTED_COMPRESSION_TYPE regardless of the
> message
> > format.
> > 2. Down-conversion of zstd-compressed records will not be supported.
> > Instead we will return UNSUPPORTED_COMPRESSION_TYPE.
> >
> > Does that sound right?
> >
> > Thanks,
> > Jason
> >
> > On Thu, Sep 6, 2018 at 1:45 AM, Magnus Edenhill <ma...@edenhill.se>
> > wrote:
> >
> > > > Ismael wrote:
> > > > Jason, that's an interesting point regarding the Java client. Do we
> > know
> > > > what clients in other languages do in these cases?
> > >
> > > librdkafka (and its bindings) passes unknown/future errors through to
> the
> > > application, the error code remains intact while
> > > the error string will be set to something like "Err-123?", which isn't
> > very
> > > helpful to the user but it at least
> > > preserves the original error code for further troubleshooting.
> > > For the producer any unknown error returned in the ProduceResponse will
> > be
> > > considered a permanent delivery failure (no retries),
> > > and for the consumer any unknown FetchResponse errors will propagate
> > > directly to the application, trigger a fetch backoff, and then
> > > continue fetching past that offset.
> > >
> > > So, from the client's perspective it is not really a problem if new
> error
> > > codes are added to older API versions.
> > >
> > > /Magnus
> > >
> > >
> > > Den tors 6 sep. 2018 kl 09:45 skrev Dongjin Lee <do...@apache.org>:
> > >
> > > > I updated the KIP page
> > > > <
> > > > https://cwiki.apache.org/confluence/display/KAFKA/KIP-
> > > 110%3A+Add+Codec+for+ZStandard+Compression
> > > > >
> > > > following the discussion here. Please take a look when you are free.
> > > > If you have any opinion, don't hesitate to give me a message.
> > > >
> > > > Best,
> > > > Dongjin
> > > >
> > > > On Fri, Aug 31, 2018 at 11:35 PM Dongjin Lee <do...@apache.org>
> > wrote:
> > > >
> > > > > I just updated the draft implementation[^1], rebasing against the
> > > latest
> > > > > trunk and implementing error routine (i.e., Error code 74 for
> > > > > UnsupportedCompressionTypeException.) Since we decided to disallow
> > all
> > > > > fetch request below version 2.1.0 for the topics specifying
> > ZStandard,
> > > I
> > > > > added an error logic only.
> > > > >
> > > > > Please have a look when you are free.
> > > > >
> > > > > Thanks,
> > > > > Dongjin
> > > > >
> > > > > [^1]: Please check the last commit here:
> > > > > https://github.com/apache/kafka/pull/2267
> > > > >
> > > > > On Thu, Aug 23, 2018, 8:55 AM Dongjin Lee <do...@apache.org>
> > wrote:
> > > > >
> > > > >> Jason,
> > > > >>
> > > > >> Great. +1 for UNSUPPORTED_COMPRESSION_TYPE.
> > > > >>
> > > > >> Best,
> > > > >> Dongjin
> > > > >>
> > > > >> On Thu, Aug 23, 2018 at 8:19 AM Jason Gustafson <
> jason@confluent.io
> > >
> > > > >> wrote:
> > > > >>
> > > > >>> Hey Dongjin,
> > > > >>>
> > > > >>> Yeah that's right. For what it's worth, librdkafka also appears
> to
> > > > handle
> > > > >>> unexpected error codes. I expect that most client implementations
> > > would
> > > > >>> either pass through the raw type or convert to an enum using
> > > something
> > > > >>> like
> > > > >>> what the java client does. Since we're expecting the client to
> fail
> > > > >>> anyway,
> > > > >>> I'm probably in favor of using the UNSUPPORTED_COMPRESSION_TYPE
> > error
> > > > >>> code.
> > > > >>>
> > > > >>> -Jason
> > > > >>>
> > > > >>> On Wed, Aug 22, 2018 at 1:46 AM, Dongjin Lee <dongjin@apache.org
> >
> > > > wrote:
> > > > >>>
> > > > >>> > Jason and Ismael,
> > > > >>> >
> > > > >>> > It seems like the only thing we need to regard if we define a
> new
> > > > error
> > > > >>> > code (i.e., UNSUPPORTED_COMPRESSION_TYPE) would be the
> > > implementation
> > > > >>> of
> > > > >>> > the other language clients, right? At least, this strategy
> causes
> > > any
> > > > >>> > problem for Java client. Do I understand correctly?
> > > > >>> >
> > > > >>> > Thanks,
> > > > >>> > Dongjin
> > > > >>> >
> > > > >>> > On Wed, Aug 22, 2018 at 5:43 PM Dongjin Lee <
> dongjin@apache.org>
> > > > >>> wrote:
> > > > >>> >
> > > > >>> > > Jason,
> > > > >>> > >
> > > > >>> > > > I think we would only use this error code when we /know/
> that
> > > > zstd
> > > > >>> was
> > > > >>> > > in use and the client doesn't support it? This is true if
> > either
> > > 1)
> > > > >>> the
> > > > >>> > > message needs down-conversion and we encounter a zstd
> > compressed
> > > > >>> message,
> > > > >>> > > or 2) if the topic is explicitly configured to use zstd.
> > > > >>> > >
> > > > >>> > > Yes, it is right. And you know, the case 1 includes 1.a) old
> > > > clients'
> > > > >>> > > request v0, v1 records or 1.b) implicit zstd, the compression
> > > type
> > > > of
> > > > >>> > > "producer" with Zstd compressed data.
> > > > >>> > >
> > > > >>> > > > However, if the compression type is set to "producer," then
> > the
> > > > >>> fetched
> > > > >>> > > data may or may not be compressed with zstd. In this case, we
> > > > return
> > > > >>> the
> > > > >>> > > data to the client and expect it to fail parsing. Is that
> > > correct?
> > > > >>> > >
> > > > >>> > > Exactly.
> > > > >>> > >
> > > > >>> > > Following your message, I reviewed the implementation of
> > > > >>> > > `KafkaApis#handleFetchRequest,` which handles the fetch
> > request.
> > > > And
> > > > >>> > found
> > > > >>> > > that the information we can use is like the following:
> > > > >>> > >
> > > > >>> > > 1. Client's fetch request version. (`versionId` variable)
> > > > >>> > > 2. Log's compression type. (`logConfig` variable)
> > > > >>> > >
> > > > >>> > > We can't detect the actual compression type of the data,
> unless
> > > we
> > > > >>> > inspect
> > > > >>> > > the `RecordBatch` included in the `Records` instance (i.e.,
> > > > >>> > > `unconvertedRecords` variable.) Since it requires some
> > > performance
> > > > >>> issue,
> > > > >>> > > it is not our option - in short, we can't be sure if given
> > chunks
> > > > of
> > > > >>> data
> > > > >>> > > are compressed with zstd or not.
> > > > >>> > >
> > > > >>> > > So, conclusion: we can return an error in the case of 1.a
> and 2
> > > > >>> easily,
> > > > >>> > > with the information above. In the case 1.b (implicit zstd),
> we
> > > can
> > > > >>> just
> > > > >>> > > return the data by do nothing special and expect it to fail
> > > > parsing.
> > > > >>> > >
> > > > >>> > > Thanks,
> > > > >>> > > Dongjin
> > > > >>> > >
> > > > >>> > > On Wed, Aug 22, 2018 at 12:02 PM Ismael Juma <
> > ismaelj@gmail.com>
> > > > >>> wrote:
> > > > >>> > >
> > > > >>> > >> Jason, that's an interesting point regarding the Java
> client.
> > Do
> > > > we
> > > > >>> know
> > > > >>> > >> what clients in other languages do in these cases?
> > > > >>> > >>
> > > > >>> > >> Ismael
> > > > >>> > >>
> > > > >>> > >> On Tue, 21 Aug 2018, 17:30 Jason Gustafson, <
> > jason@confluent.io
> > > >
> > > > >>> wrote:
> > > > >>> > >>
> > > > >>> > >> > Hi Dongjin,
> > > > >>> > >> >
> > > > >>> > >> > One of the complications is that old versions of the API
> > will
> > > > not
> > > > >>> > >> expect a
> > > > >>> > >> > new error code. However, since we expect this to be a
> fatal
> > > > error
> > > > >>> > anyway
> > > > >>> > >> > for old clients, it may still be more useful to return the
> > > > correct
> > > > >>> > error
> > > > >>> > >> > code. For example, the Kafka clients use the following
> code
> > to
> > > > >>> convert
> > > > >>> > >> the
> > > > >>> > >> > error code:
> > > > >>> > >> >
> > > > >>> > >> >     public static Errors forCode(short code) {
> > > > >>> > >> >         Errors error = codeToError.get(code);
> > > > >>> > >> >         if (error != null) {
> > > > >>> > >> >             return error;
> > > > >>> > >> >         } else {
> > > > >>> > >> >             log.warn("Unexpected error code: {}.", code);
> > > > >>> > >> >             return UNKNOWN_SERVER_ERROR;
> > > > >>> > >> >         }
> > > > >>> > >> >     }
> > > > >>> > >> >
> > > > >>> > >> > If we return an unsupported error code, it will be
> converted
> > > to
> > > > an
> > > > >>> > >> UNKNOWN
> > > > >>> > >> > error, but at least we will get the message in the log
> with
> > > the
> > > > >>> > correct
> > > > >>> > >> > code. That seems preferable to returning a misleading
> error
> > > > code.
> > > > >>> So I
> > > > >>> > >> > wonder if we can use the new UNSUPPORTED_COMPRESSION_TYPE
> > > error
> > > > >>> even
> > > > >>> > for
> > > > >>> > >> > older versions.
> > > > >>> > >> >
> > > > >>> > >> > Also, one question just to check my understanding. I think
> > we
> > > > >>> would
> > > > >>> > only
> > > > >>> > >> > use this error code when we /know/ that zstd was in use
> and
> > > the
> > > > >>> client
> > > > >>> > >> > doesn't support it? This is true if either 1) the message
> > > needs
> > > > >>> > >> > down-conversion and we encounter a zstd compressed
> message,
> > or
> > > > 2)
> > > > >>> if
> > > > >>> > the
> > > > >>> > >> > topic is explicitly configured to use zstd. However, if
> the
> > > > >>> > compression
> > > > >>> > >> > type is set to "producer," then the fetched data may or
> may
> > > not
> > > > be
> > > > >>> > >> > compressed with zstd. In this case, we return the data to
> > the
> > > > >>> client
> > > > >>> > and
> > > > >>> > >> > expect it to fail parsing. Is that correct?
> > > > >>> > >> >
> > > > >>> > >> > Thanks,
> > > > >>> > >> > Jason
> > > > >>> > >> >
> > > > >>> > >> >
> > > > >>> > >> >
> > > > >>> > >> > On Tue, Aug 21, 2018 at 9:08 AM, Dongjin Lee <
> > > > dongjin@apache.org>
> > > > >>> > >> wrote:
> > > > >>> > >> >
> > > > >>> > >> > > Ismael, Jason and all,
> > > > >>> > >> > >
> > > > >>> > >> > > I rewrote the backward compatibility strategy & its
> > > > alternatives
> > > > >>> > like
> > > > >>> > >> > > following, based on Ismael & Jason's comments. Since it
> is
> > > not
> > > > >>> > >> updated to
> > > > >>> > >> > > the wiki yet, don't hesitate to give me a message if you
> > > have
> > > > >>> any
> > > > >>> > >> opinion
> > > > >>> > >> > > on it.
> > > > >>> > >> > >
> > > > >>> > >> > > ```
> > > > >>> > >> > > *Backward Compatibility*
> > > > >>> > >> > >
> > > > >>> > >> > > We need to establish some backward-compatibility
> strategy
> > > for
> > > > >>> the
> > > > >>> > >> case an
> > > > >>> > >> > > old client subscribes a topic using ZStandard implicitly
> > > > (i.e.,
> > > > >>> > >> > > 'compression.type' configuration of given topic is
> > > 'producer'
> > > > >>> and
> > > > >>> > the
> > > > >>> > >> > > producer compressed the records with ZStandard). We have
> > the
> > > > >>> > following
> > > > >>> > >> > > options for this situation:
> > > > >>> > >> > >
> > > > >>> > >> > > *A. Support ZStandard to the old clients which can
> > > understand
> > > > >>> v0, v1
> > > > >>> > >> > > messages only.*
> > > > >>> > >> > >
> > > > >>> > >> > > This strategy necessarily requires the down-conversion
> of
> > v2
> > > > >>> message
> > > > >>> > >> > > compressed with Zstandard into v0 or v1 messages, which
> > > means
> > > > a
> > > > >>> > >> > > considerable performance degradation. So we rejected
> this
> > > > >>> strategy.
> > > > >>> > >> > >
> > > > >>> > >> > > *B. Bump the API version and support only v2-available
> > > > clients*
> > > > >>> > >> > >
> > > > >>> > >> > > With this approach, we can message the old clients that
> > they
> > > > >>> are old
> > > > >>> > >> and
> > > > >>> > >> > > should be upgraded. However, there are still several
> > options
> > > > >>> for the
> > > > >>> > >> > Error
> > > > >>> > >> > > code.
> > > > >>> > >> > >
> > > > >>> > >> > > *B.1. INVALID_REQUEST (42)*
> > > > >>> > >> > >
> > > > >>> > >> > > This option gives the client so little information; the
> > user
> > > > >>> can be
> > > > >>> > >> > > confused about why the client worked correctly in the
> past
> > > > >>> suddenly
> > > > >>> > >> > > encounters a problem. So we rejected this strategy.
> > > > >>> > >> > >
> > > > >>> > >> > > *B.2. CORRUPT_MESSAGE (2)*
> > > > >>> > >> > >
> > > > >>> > >> > > This option gives inaccurate information; the user can
> be
> > > > >>> surprised
> > > > >>> > >> and
> > > > >>> > >> > > misunderstand that the log files are broken in some way.
> > So
> > > we
> > > > >>> > >> rejected
> > > > >>> > >> > > this strategy.
> > > > >>> > >> > >
> > > > >>> > >> > > *B.3 UNSUPPORTED_FOR_MESSAGE_FORMAT (43)*
> > > > >>> > >> > >
> > > > >>> > >> > > The advantage of this approach is we don't need to
> define
> > a
> > > > new
> > > > >>> > error
> > > > >>> > >> > code;
> > > > >>> > >> > > we can reuse it and that's all.
> > > > >>> > >> > >
> > > > >>> > >> > > The disadvantage of this approach is that it is also a
> > > little
> > > > >>> bit
> > > > >>> > >> vague;
> > > > >>> > >> > > This error code is defined as a work for KIP-98[^1] and
> > now
> > > > >>> returned
> > > > >>> > >> in
> > > > >>> > >> > the
> > > > >>> > >> > > transaction error.
> > > > >>> > >> > >
> > > > >>> > >> > > *B.4. UNSUPPORTED_COMPRESSION_TYPE (new)*
> > > > >>> > >> > >
> > > > >>> > >> > > The advantage of this approach is that it is clear and
> > > > provides
> > > > >>> an
> > > > >>> > >> exact
> > > > >>> > >> > > description. The disadvantage is we need to add a new
> > error
> > > > >>> code.
> > > > >>> > >> > > ```
> > > > >>> > >> > >
> > > > >>> > >> > > *It seems like what we need to choose is now so clear:
> > > > >>> > >> > > UNSUPPORTED_FOR_MESSAGE_FORMAT (B.3) or
> > > > >>> UNSUPPORTED_COMPRESSION_TYPE
> > > > >>> > >> > > (B.4).*
> > > > >>> > >> > > The first one doesn't need a new error message but the
> > > latter
> > > > is
> > > > >>> > more
> > > > >>> > >> > > explicit. Which one do you prefer? Since all of you have
> > > much
> > > > >>> more
> > > > >>> > >> > > experience and knowledge than me, I will follow your
> > > decision.
> > > > >>> The
> > > > >>> > >> wiki
> > > > >>> > >> > > page will be updated following the decision also.
> > > > >>> > >> > >
> > > > >>> > >> > > Best,
> > > > >>> > >> > > Dongjin
> > > > >>> > >> > >
> > > > >>> > >> > > [^1]: https://issues.apache.org/jira/browse/KAFKA-4990
> > > > >>> > >> > >
> > > > >>> > >> > > On Sun, Aug 19, 2018 at 4:58 AM Ismael Juma <
> > > > ismaelj@gmail.com>
> > > > >>> > >> wrote:
> > > > >>> > >> > >
> > > > >>> > >> > > > Sounds reasonable to me.
> > > > >>> > >> > > >
> > > > >>> > >> > > > Ismael
> > > > >>> > >> > > >
> > > > >>> > >> > > > On Sat, 18 Aug 2018, 12:20 Jason Gustafson, <
> > > > >>> jason@confluent.io>
> > > > >>> > >> > wrote:
> > > > >>> > >> > > >
> > > > >>> > >> > > > > Hey Ismael,
> > > > >>> > >> > > > >
> > > > >>> > >> > > > > Your summary looks good to me. I think it might also
> > be
> > > a
> > > > >>> good
> > > > >>> > >> idea
> > > > >>> > >> > to
> > > > >>> > >> > > > add
> > > > >>> > >> > > > > a new UNSUPPORTED_COMPRESSION_TYPE error code to go
> > > along
> > > > >>> with
> > > > >>> > the
> > > > >>> > >> > > > version
> > > > >>> > >> > > > > bumps. We won't be able to use it for old api
> versions
> > > > >>> since the
> > > > >>> > >> > > clients
> > > > >>> > >> > > > > will not understand it, but we can use it going
> > forward
> > > so
> > > > >>> that
> > > > >>> > >> we're
> > > > >>> > >> > > not
> > > > >>> > >> > > > > stuck in a similar situation with a new message
> format
> > > and
> > > > >>> a new
> > > > >>> > >> > codec
> > > > >>> > >> > > to
> > > > >>> > >> > > > > support. Another option is to use
> > > UNSUPPORTED_FOR_MESSAGE_
> > > > >>> > FORMAT,
> > > > >>> > >> but
> > > > >>> > >> > > it
> > > > >>> > >> > > > is
> > > > >>> > >> > > > > not as explicit.
> > > > >>> > >> > > > >
> > > > >>> > >> > > > > -Jason
> > > > >>> > >> > > > >
> > > > >>> > >> > > > > On Fri, Aug 17, 2018 at 5:19 PM, Ismael Juma <
> > > > >>> ismael@juma.me.uk
> > > > >>> > >
> > > > >>> > >> > > wrote:
> > > > >>> > >> > > > >
> > > > >>> > >> > > > > > Hi Dongjin and Jason,
> > > > >>> > >> > > > > >
> > > > >>> > >> > > > > > I would agree. My summary:
> > > > >>> > >> > > > > >
> > > > >>> > >> > > > > > 1. Support zstd with message format 2 only.
> > > > >>> > >> > > > > > 2. Bump produce and fetch request versions.
> > > > >>> > >> > > > > > 3. Provide broker errors whenever possible based
> on
> > > the
> > > > >>> > request
> > > > >>> > >> > > version
> > > > >>> > >> > > > > and
> > > > >>> > >> > > > > > rely on clients for the cases where the broker
> can't
> > > > >>> validate
> > > > >>> > >> > > > efficiently
> > > > >>> > >> > > > > > (example message format 2 consumer that supports
> the
> > > > >>> latest
> > > > >>> > >> fetch
> > > > >>> > >> > > > version
> > > > >>> > >> > > > > > but doesn't support zstd).
> > > > >>> > >> > > > > >
> > > > >>> > >> > > > > > If there's general agreement on this, I suggest we
> > > > update
> > > > >>> the
> > > > >>> > >> KIP
> > > > >>> > >> > to
> > > > >>> > >> > > > > state
> > > > >>> > >> > > > > > the proposal and to move the rejected options to
> its
> > > own
> > > > >>> > >> section.
> > > > >>> > >> > And
> > > > >>> > >> > > > > then
> > > > >>> > >> > > > > > start a vote!
> > > > >>> > >> > > > > >
> > > > >>> > >> > > > > > Ismael
> > > > >>> > >> > > > > >
> > > > >>> > >> > > > > > On Fri, Aug 17, 2018 at 4:00 PM Jason Gustafson <
> > > > >>> > >> > jason@confluent.io>
> > > > >>> > >> > > > > > wrote:
> > > > >>> > >> > > > > >
> > > > >>> > >> > > > > > > Hi Dongjin,
> > > > >>> > >> > > > > > >
> > > > >>> > >> > > > > > > Yes, that's a good summary. For clients which
> > > support
> > > > >>> v2,
> > > > >>> > the
> > > > >>> > >> > > client
> > > > >>> > >> > > > > can
> > > > >>> > >> > > > > > > parse the message format and hopefully raise a
> > > useful
> > > > >>> error
> > > > >>> > >> > message
> > > > >>> > >> > > > > > > indicating the unsupported compression type. For
> > > older
> > > > >>> > >> clients,
> > > > >>> > >> > our
> > > > >>> > >> > > > > > options
> > > > >>> > >> > > > > > > are probably (1) to down-convert to the old
> format
> > > > >>> using no
> > > > >>> > >> > > > compression
> > > > >>> > >> > > > > > > type, or (2) to return an error code. I'm
> leaning
> > > > >>> toward the
> > > > >>> > >> > latter
> > > > >>> > >> > > > as
> > > > >>> > >> > > > > > the
> > > > >>> > >> > > > > > > simpler solution, but the challenge is finding a
> > > good
> > > > >>> error
> > > > >>> > >> code.
> > > > >>> > >> > > Two
> > > > >>> > >> > > > > > > possibilities might be INVALID_REQUEST or
> > > > >>> CORRUPT_MESSAGE.
> > > > >>> > The
> > > > >>> > >> > > > downside
> > > > >>> > >> > > > > > is
> > > > >>> > >> > > > > > > that old clients probably won't get a helpful
> > > message.
> > > > >>> > >> However,
> > > > >>> > >> > at
> > > > >>> > >> > > > > least
> > > > >>> > >> > > > > > > the behavior will be consistent in the sense
> that
> > > all
> > > > >>> > clients
> > > > >>> > >> > will
> > > > >>> > >> > > > fail
> > > > >>> > >> > > > > > if
> > > > >>> > >> > > > > > > they do not support zstandard.
> > > > >>> > >> > > > > > >
> > > > >>> > >> > > > > > > What do you think?
> > > > >>> > >> > > > > > >
> > > > >>> > >> > > > > > > Thanks,
> > > > >>> > >> > > > > > > Jason
> > > > >>> > >> > > > > > >
> > > > >>> > >> > > > > > > On Fri, Aug 17, 2018 at 8:08 AM, Dongjin Lee <
> > > > >>> > >> dongjin@apache.org
> > > > >>> > >> > >
> > > > >>> > >> > > > > wrote:
> > > > >>> > >> > > > > > >
> > > > >>> > >> > > > > > > > Thanks Jason, I reviewed the down-converting
> > logic
> > > > >>> > following
> > > > >>> > >> > your
> > > > >>> > >> > > > > > > > explanation.[^1] You mean the following
> > routines,
> > > > >>> right?
> > > > >>> > >> > > > > > > >
> > > > >>> > >> > > > > > > > -
> > > > >>> > >> > > > > > > >
> > > > https://github.com/apache/kafka/blob/trunk/core/src/
> > > > >>> > >> > > > > > > > main/scala/kafka/server/KafkaApis.scala#L534
> > > > >>> > >> > > > > > > > -
> > > > >>> > >> > > > > > > >
> > > > >>> https://github.com/apache/kafka/blob/trunk/clients/src/
> > > > >>> > >> > > > > > > > main/java/org/apache/kafka/common/record/
> > > > >>> > >> > > LazyDownConversionRecords.
> > > > >>> > >> > > > > > > > java#L165
> > > > >>> > >> > > > > > > > -
> > > > >>> > >> > > > > > > >
> > > > >>> https://github.com/apache/kafka/blob/trunk/clients/src/
> > > > >>> > >> > > > > > > >
> > > > >>> > >>
> main/java/org/apache/kafka/common/record/RecordsUtil.java#L40
> > > > >>> > >> > > > > > > >
> > > > >>> > >> > > > > > > > It seems like your stance is like following:
> > > > >>> > >> > > > > > > >
> > > > >>> > >> > > > > > > > 1. In principle, Kafka does not change the
> > > > compression
> > > > >>> > codec
> > > > >>> > >> > when
> > > > >>> > >> > > > > > > > down-converting, since it requires inspecting
> > the
> > > > >>> fetched
> > > > >>> > >> data,
> > > > >>> > >> > > > which
> > > > >>> > >> > > > > > is
> > > > >>> > >> > > > > > > > expensive.
> > > > >>> > >> > > > > > > > 2. However, there are some cases the fetched
> > data
> > > is
> > > > >>> > >> inspected
> > > > >>> > >> > > > > anyway.
> > > > >>> > >> > > > > > In
> > > > >>> > >> > > > > > > > this case, we can provide compression
> conversion
> > > > from
> > > > >>> > >> Zstandard
> > > > >>> > >> > > to
> > > > >>> > >> > > > > > > > classical ones[^2].
> > > > >>> > >> > > > > > > >
> > > > >>> > >> > > > > > > > And from what I understand, the cases where
> the
> > > > client
> > > > >>> > >> without
> > > > >>> > >> > > > > > ZStandard
> > > > >>> > >> > > > > > > > support receives ZStandard compressed records
> > can
> > > be
> > > > >>> > >> organized
> > > > >>> > >> > > into
> > > > >>> > >> > > > > two
> > > > >>> > >> > > > > > > > cases:
> > > > >>> > >> > > > > > > >
> > > > >>> > >> > > > > > > > a. The 'compression.type' configuration of
> given
> > > > >>> topic is
> > > > >>> > >> > > > 'producer'
> > > > >>> > >> > > > > > and
> > > > >>> > >> > > > > > > > the producer compressed the records with
> > > ZStandard.
> > > > >>> (that
> > > > >>> > >> is,
> > > > >>> > >> > > using
> > > > >>> > >> > > > > > > > ZStandard implicitly.)
> > > > >>> > >> > > > > > > > b.  The 'compression.type' configuration of
> > given
> > > > >>> topic is
> > > > >>> > >> > > 'zstd';
> > > > >>> > >> > > > > that
> > > > >>> > >> > > > > > > is,
> > > > >>> > >> > > > > > > > using ZStandard explicitly.
> > > > >>> > >> > > > > > > >
> > > > >>> > >> > > > > > > > As you stated, we don't have to handle the
> case
> > b
> > > > >>> > specially.
> > > > >>> > >> > So,
> > > > >>> > >> > > It
> > > > >>> > >> > > > > > seems
> > > > >>> > >> > > > > > > > like we can narrow the focus of the problem by
> > > > joining
> > > > >>> > case
> > > > >>> > >> 1
> > > > >>> > >> > and
> > > > >>> > >> > > > > case
> > > > >>> > >> > > > > > b
> > > > >>> > >> > > > > > > > like the following:
> > > > >>> > >> > > > > > > >
> > > > >>> > >> > > > > > > > > Given the topic with 'producer' as its
> > > > >>> > 'compression.type'
> > > > >>> > >> > > > > > > configuration,
> > > > >>> > >> > > > > > > > ZStandard compressed records and old client
> > > without
> > > > >>> > >> ZStandard,
> > > > >>> > >> > is
> > > > >>> > >> > > > > there
> > > > >>> > >> > > > > > > any
> > > > >>> > >> > > > > > > > case we need to inspect the records and can
> > change
> > > > the
> > > > >>> > >> > > compression
> > > > >>> > >> > > > > > type?
> > > > >>> > >> > > > > > > If
> > > > >>> > >> > > > > > > > so, can we provide compression type
> converting?
> > > > >>> > >> > > > > > > >
> > > > >>> > >> > > > > > > > Do I understand correctly?
> > > > >>> > >> > > > > > > >
> > > > >>> > >> > > > > > > > Best,
> > > > >>> > >> > > > > > > > Dongjin
> > > > >>> > >> > > > > > > >
> > > > >>> > >> > > > > > > > [^1]: I'm sorry, I found that I was a little
> bit
> > > > >>> > >> > misunderstanding
> > > > >>> > >> > > > how
> > > > >>> > >> > > > > > API
> > > > >>> > >> > > > > > > > version works, after reviewing the downconvert
> > > logic
> > > > >>> & the
> > > > >>> > >> > > protocol
> > > > >>> > >> > > > > > > > documentation <
> > https://kafka.apache.org/protocol
> > > >.
> > > > >>> > >> > > > > > > > [^2]: None, Gzip, Snappy, Lz4.
> > > > >>> > >> > > > > > > >
> > > > >>> > >> > > > > > > > On Tue, Aug 14, 2018 at 2:16 AM Jason
> Gustafson
> > <
> > > > >>> > >> > > > jason@confluent.io>
> > > > >>> > >> > > > > > > > wrote:
> > > > >>> > >> > > > > > > >
> > > > >>> > >> > > > > > > > > >
> > > > >>> > >> > > > > > > > > > But in my opinion, since the client will
> > fail
> > > > >>> with the
> > > > >>> > >> API
> > > > >>> > >> > > > > version,
> > > > >>> > >> > > > > > > so
> > > > >>> > >> > > > > > > > we
> > > > >>> > >> > > > > > > > > > don't need to down-convert the messages
> > > anyway.
> > > > >>> Isn't
> > > > >>> > >> it?
> > > > >>> > >> > > So, I
> > > > >>> > >> > > > > > think
> > > > >>> > >> > > > > > > > we
> > > > >>> > >> > > > > > > > > > don't care about this case. (I'm sorry, I
> am
> > > not
> > > > >>> > >> familiar
> > > > >>> > >> > > with
> > > > >>> > >> > > > > > > > > down-convert
> > > > >>> > >> > > > > > > > > > logic.)
> > > > >>> > >> > > > > > > > >
> > > > >>> > >> > > > > > > > >
> > > > >>> > >> > > > > > > > > Currently the broker down-converts
> > automatically
> > > > >>> when it
> > > > >>> > >> > > receives
> > > > >>> > >> > > > > an
> > > > >>> > >> > > > > > > old
> > > > >>> > >> > > > > > > > > version of the fetch request (a version
> which
> > is
> > > > >>> known
> > > > >>> > to
> > > > >>> > >> > > predate
> > > > >>> > >> > > > > the
> > > > >>> > >> > > > > > > > > message format in use). Typically when
> > > > >>> down-converting
> > > > >>> > the
> > > > >>> > >> > > > message
> > > > >>> > >> > > > > > > > format,
> > > > >>> > >> > > > > > > > > we use the same compression type, but there
> is
> > > not
> > > > >>> much
> > > > >>> > >> point
> > > > >>> > >> > > in
> > > > >>> > >> > > > > > doing
> > > > >>> > >> > > > > > > so
> > > > >>> > >> > > > > > > > > when we know the client doesn't support it.
> So
> > > if
> > > > >>> > >> zstandard
> > > > >>> > >> > is
> > > > >>> > >> > > in
> > > > >>> > >> > > > > > use,
> > > > >>> > >> > > > > > > > and
> > > > >>> > >> > > > > > > > > we have to down-convert anyway, then we can
> > > choose
> > > > >>> to
> > > > >>> > use
> > > > >>> > >> a
> > > > >>> > >> > > > > different
> > > > >>> > >> > > > > > > > > compression type or no compression type.
> > > > >>> > >> > > > > > > > >
> > > > >>> > >> > > > > > > > > From my perspective, there is no significant
> > > > >>> downside to
> > > > >>> > >> > > bumping
> > > > >>> > >> > > > > the
> > > > >>> > >> > > > > > > > > protocol version and it has several
> potential
> > > > >>> benefits.
> > > > >>> > >> > Version
> > > > >>> > >> > > > > bumps
> > > > >>> > >> > > > > > > are
> > > > >>> > >> > > > > > > > > cheap. The main question mark in my mind is
> > > about
> > > > >>> > >> > > > down-conversion.
> > > > >>> > >> > > > > > > > Figuring
> > > > >>> > >> > > > > > > > > out whether down-conversion is needed is
> hard
> > > > >>> generally
> > > > >>> > >> > without
> > > > >>> > >> > > > > > > > inspecting
> > > > >>> > >> > > > > > > > > the fetched data, which is expensive. I
> think
> > we
> > > > >>> agree
> > > > >>> > in
> > > > >>> > >> > > > principle
> > > > >>> > >> > > > > > > that
> > > > >>> > >> > > > > > > > we
> > > > >>> > >> > > > > > > > > do not want to have to pay this cost
> generally
> > > and
> > > > >>> > prefer
> > > > >>> > >> the
> > > > >>> > >> > > > > clients
> > > > >>> > >> > > > > > > to
> > > > >>> > >> > > > > > > > > fail when they see an unhandled compression
> > > type.
> > > > >>> The
> > > > >>> > >> point I
> > > > >>> > >> > > was
> > > > >>> > >> > > > > > > making
> > > > >>> > >> > > > > > > > is
> > > > >>> > >> > > > > > > > > that there are some cases where we are
> either
> > > > >>> inspecting
> > > > >>> > >> the
> > > > >>> > >> > > data
> > > > >>> > >> > > > > > > anyway
> > > > >>> > >> > > > > > > > > (because we have to down-convert the message
> > > > >>> format), or
> > > > >>> > >> we
> > > > >>> > >> > > have
> > > > >>> > >> > > > an
> > > > >>> > >> > > > > > > easy
> > > > >>> > >> > > > > > > > > way to tell whether zstandard is in use (the
> > > topic
> > > > >>> has
> > > > >>> > it
> > > > >>> > >> > > > > configured
> > > > >>> > >> > > > > > > > > explicitly). In the latter case, we don't
> have
> > > to
> > > > >>> handle
> > > > >>> > >> it
> > > > >>> > >> > > > > > specially.
> > > > >>> > >> > > > > > > > But
> > > > >>> > >> > > > > > > > > we do have to decide how we will handle
> > > > >>> down-conversion
> > > > >>> > to
> > > > >>> > >> > > older
> > > > >>> > >> > > > > > > formats.
> > > > >>> > >> > > > > > > > >
> > > > >>> > >> > > > > > > > > -Jason
> > > > >>> > >> > > > > > > > >
> > > > >>> > >> > > > > > > > > On Sun, Aug 12, 2018 at 5:15 PM, Dongjin
> Lee <
> > > > >>> > >> > > dongjin@apache.org
> > > > >>> > >> > > > >
> > > > >>> > >> > > > > > > wrote:
> > > > >>> > >> > > > > > > > >
> > > > >>> > >> > > > > > > > > > Colin and Jason,
> > > > >>> > >> > > > > > > > > >
> > > > >>> > >> > > > > > > > > > Thanks for your opinions. In summarizing,
> > the
> > > > >>> Pros and
> > > > >>> > >> Cons
> > > > >>> > >> > > of
> > > > >>> > >> > > > > > > bumping
> > > > >>> > >> > > > > > > > > > fetch API version are:
> > > > >>> > >> > > > > > > > > >
> > > > >>> > >> > > > > > > > > > Cons:
> > > > >>> > >> > > > > > > > > >
> > > > >>> > >> > > > > > > > > > - The Broker can't know whether a given
> > > message
> > > > >>> batch
> > > > >>> > is
> > > > >>> > >> > > > > compressed
> > > > >>> > >> > > > > > > > with
> > > > >>> > >> > > > > > > > > > zstd or not.
> > > > >>> > >> > > > > > > > > > - Need some additional logic for the topic
> > > > >>> explicitly
> > > > >>> > >> > > > configured
> > > > >>> > >> > > > > to
> > > > >>> > >> > > > > > > use
> > > > >>> > >> > > > > > > > > > zstd.
> > > > >>> > >> > > > > > > > > >
> > > > >>> > >> > > > > > > > > > Pros:
> > > > >>> > >> > > > > > > > > >
> > > > >>> > >> > > > > > > > > > - The broker doesn't need to conduct
> > expensive
> > > > >>> > >> > > down-conversion.
> > > > >>> > >> > > > > > > > > > - Can message the users to update their
> > > client.
> > > > >>> > >> > > > > > > > > >
> > > > >>> > >> > > > > > > > > > So, opinions for the
> backward-compatibility
> > > > >>> policy by
> > > > >>> > >> far:
> > > > >>> > >> > > > > > > > > >
> > > > >>> > >> > > > > > > > > > - A: bump the API version - +2 (Colin,
> > Jason)
> > > > >>> > >> > > > > > > > > > - B: leave unchanged - +1 (Viktor)
> > > > >>> > >> > > > > > > > > >
> > > > >>> > >> > > > > > > > > > Here are my additional comments:
> > > > >>> > >> > > > > > > > > >
> > > > >>> > >> > > > > > > > > > @Colin
> > > > >>> > >> > > > > > > > > >
> > > > >>> > >> > > > > > > > > > I greatly appreciate your response. In the
> > > case
> > > > >>> of the
> > > > >>> > >> > > > dictionary
> > > > >>> > >> > > > > > > > > support,
> > > > >>> > >> > > > > > > > > > of course, this issue should be addressed
> > > later
> > > > >>> so we
> > > > >>> > >> don't
> > > > >>> > >> > > > need
> > > > >>> > >> > > > > it
> > > > >>> > >> > > > > > > in
> > > > >>> > >> > > > > > > > > the
> > > > >>> > >> > > > > > > > > > first version. You are right - it is not
> > late
> > > to
> > > > >>> try
> > > > >>> > it
> > > > >>> > >> > after
> > > > >>> > >> > > > > some
> > > > >>> > >> > > > > > > > > > benchmarks. What I mean is, we should keep
> > in
> > > > >>> mind on
> > > > >>> > >> that
> > > > >>> > >> > > > > > potential
> > > > >>> > >> > > > > > > > > > feature.
> > > > >>> > >> > > > > > > > > >
> > > > >>> > >> > > > > > > > > > @Jason
> > > > >>> > >> > > > > > > > > >
> > > > >>> > >> > > > > > > > > > You wrote,
> > > > >>> > >> > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > Similarly, if we have to down-convert
> > anyway
> > > > >>> because
> > > > >>> > >> the
> > > > >>> > >> > > > client
> > > > >>> > >> > > > > > > does
> > > > >>> > >> > > > > > > > > not
> > > > >>> > >> > > > > > > > > > understand the message format, then we
> could
> > > > also
> > > > >>> use
> > > > >>> > a
> > > > >>> > >> > > > different
> > > > >>> > >> > > > > > > > > > compression type.
> > > > >>> > >> > > > > > > > > >
> > > > >>> > >> > > > > > > > > > But in my opinion, since the client will
> > fail
> > > > >>> with the
> > > > >>> > >> API
> > > > >>> > >> > > > > version,
> > > > >>> > >> > > > > > > so
> > > > >>> > >> > > > > > > > we
> > > > >>> > >> > > > > > > > > > don't need to down-convert the messages
> > > anyway.
> > > > >>> Isn't
> > > > >>> > >> it?
> > > > >>> > >> > > So, I
> > > > >>> > >> > > > > > think
> > > > >>> > >> > > > > > > > we
> > > > >>> > >> > > > > > > > > > don't care about this case. (I'm sorry, I
> am
> > > not
> > > > >>> > >> familiar
> > > > >>> > >> > > with
> > > > >>> > >> > > > > > > > > down-convert
> > > > >>> > >> > > > > > > > > > logic.)
> > > > >>> > >> > > > > > > > > >
> > > > >>> > >> > > > > > > > > > Please give more opinions. Thanks!
> > > > >>> > >> > > > > > > > > >
> > > > >>> > >> > > > > > > > > > - Dongjin
> > > > >>> > >> > > > > > > > > >
> > > > >>> > >> > > > > > > > > >
> > > > >>> > >> > > > > > > > > > On Wed, Aug 8, 2018 at 6:41 AM Jason
> > > Gustafson <
> > > > >>> > >> > > > > jason@confluent.io
> > > > >>> > >> > > > > > >
> > > > >>> > >> > > > > > > > > wrote:
> > > > >>> > >> > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > Hey Colin,
> > > > >>> > >> > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > The problem for the fetch API is that
> the
> > > > broker
> > > > >>> > does
> > > > >>> > >> not
> > > > >>> > >> > > > > > generally
> > > > >>> > >> > > > > > > > > know
> > > > >>> > >> > > > > > > > > > if
> > > > >>> > >> > > > > > > > > > > a batch was compressed with zstd unless
> it
> > > > >>> parses
> > > > >>> > it.
> > > > >>> > >> I
> > > > >>> > >> > > think
> > > > >>> > >> > > > > the
> > > > >>> > >> > > > > > > > goal
> > > > >>> > >> > > > > > > > > > here
> > > > >>> > >> > > > > > > > > > > is to avoid the expensive
> down-conversion
> > > that
> > > > >>> is
> > > > >>> > >> needed
> > > > >>> > >> > to
> > > > >>> > >> > > > > > ensure
> > > > >>> > >> > > > > > > > > > > compatibility because it is only
> necessary
> > > if
> > > > >>> zstd
> > > > >>> > is
> > > > >>> > >> > > > actually
> > > > >>> > >> > > > > in
> > > > >>> > >> > > > > > > > use.
> > > > >>> > >> > > > > > > > > > But
> > > > >>> > >> > > > > > > > > > > as long as old clients can parse the
> > message
> > > > >>> format,
> > > > >>> > >> they
> > > > >>> > >> > > > > should
> > > > >>> > >> > > > > > > get
> > > > >>> > >> > > > > > > > a
> > > > >>> > >> > > > > > > > > > > reasonable error if they see an
> > unsupported
> > > > >>> > >> compression
> > > > >>> > >> > > type
> > > > >>> > >> > > > in
> > > > >>> > >> > > > > > the
> > > > >>> > >> > > > > > > > > > > attributes. Basically the onus is on
> users
> > > to
> > > > >>> ensure
> > > > >>> > >> that
> > > > >>> > >> > > > their
> > > > >>> > >> > > > > > > > > consumers
> > > > >>> > >> > > > > > > > > > > have been updated prior to using zstd.
> It
> > > > seems
> > > > >>> > like a
> > > > >>> > >> > > > > reasonable
> > > > >>> > >> > > > > > > > > > tradeoff
> > > > >>> > >> > > > > > > > > > > to me. There are a couple cases that
> might
> > > be
> > > > >>> worth
> > > > >>> > >> > > thinking
> > > > >>> > >> > > > > > > through:
> > > > >>> > >> > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > 1. If a topic is explicitly configured
> to
> > > use
> > > > >>> zstd,
> > > > >>> > >> then
> > > > >>> > >> > we
> > > > >>> > >> > > > > don't
> > > > >>> > >> > > > > > > > need
> > > > >>> > >> > > > > > > > > to
> > > > >>> > >> > > > > > > > > > > check the fetched data for the
> compression
> > > > type
> > > > >>> to
> > > > >>> > >> know
> > > > >>> > >> > if
> > > > >>> > >> > > we
> > > > >>> > >> > > > > > need
> > > > >>> > >> > > > > > > > > > > down-conversion. If we did bump the
> Fetch
> > > API
> > > > >>> > version,
> > > > >>> > >> > then
> > > > >>> > >> > > > we
> > > > >>> > >> > > > > > > could
> > > > >>> > >> > > > > > > > > > handle
> > > > >>> > >> > > > > > > > > > > this case by either down-converting
> using
> > a
> > > > >>> > different
> > > > >>> > >> > > > > compression
> > > > >>> > >> > > > > > > > type
> > > > >>> > >> > > > > > > > > or
> > > > >>> > >> > > > > > > > > > > returning an error.
> > > > >>> > >> > > > > > > > > > > 2. Similarly, if we have to down-convert
> > > > anyway
> > > > >>> > >> because
> > > > >>> > >> > the
> > > > >>> > >> > > > > > client
> > > > >>> > >> > > > > > > > does
> > > > >>> > >> > > > > > > > > > not
> > > > >>> > >> > > > > > > > > > > understand the message format, then we
> > could
> > > > >>> also
> > > > >>> > use
> > > > >>> > >> a
> > > > >>> > >> > > > > different
> > > > >>> > >> > > > > > > > > > > compression type.
> > > > >>> > >> > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > For the produce API, I think it's
> > reasonable
> > > > to
> > > > >>> bump
> > > > >>> > >> the
> > > > >>> > >> > > api
> > > > >>> > >> > > > > > > version.
> > > > >>> > >> > > > > > > > > > This
> > > > >>> > >> > > > > > > > > > > can be used by clients to check whether
> a
> > > > broker
> > > > >>> > >> supports
> > > > >>> > >> > > > zstd.
> > > > >>> > >> > > > > > For
> > > > >>> > >> > > > > > > > > > > example, we might support a list of
> > > preferred
> > > > >>> > >> compression
> > > > >>> > >> > > > types
> > > > >>> > >> > > > > > in
> > > > >>> > >> > > > > > > > the
> > > > >>> > >> > > > > > > > > > > producer and we could use the broker to
> > > detect
> > > > >>> which
> > > > >>> > >> > > version
> > > > >>> > >> > > > to
> > > > >>> > >> > > > > > > use.
> > > > >>> > >> > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > -Jason
> > > > >>> > >> > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > On Tue, Aug 7, 2018 at 1:32 PM, Colin
> > > McCabe <
> > > > >>> > >> > > > > cmccabe@apache.org
> > > > >>> > >> > > > > > >
> > > > >>> > >> > > > > > > > > wrote:
> > > > >>> > >> > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > > Thanks for bumping this, Dongjin.
> ZStd
> > > is a
> > > > >>> good
> > > > >>> > >> > > > compression
> > > > >>> > >> > > > > > > codec
> > > > >>> > >> > > > > > > > > > and I
> > > > >>> > >> > > > > > > > > > > > hope we can get this support in soon!
> > > > >>> > >> > > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > > I would say we can just bump the API
> > > version
> > > > >>> to
> > > > >>> > >> > indicate
> > > > >>> > >> > > > that
> > > > >>> > >> > > > > > > ZStd
> > > > >>> > >> > > > > > > > > > > support
> > > > >>> > >> > > > > > > > > > > > is expected in new clients.  We
> probably
> > > > need
> > > > >>> some
> > > > >>> > >> way
> > > > >>> > >> > of
> > > > >>> > >> > > > > > > > indicating
> > > > >>> > >> > > > > > > > > to
> > > > >>> > >> > > > > > > > > > > the
> > > > >>> > >> > > > > > > > > > > > older clients that they can't consume
> > the
> > > > >>> > >> partitions,
> > > > >>> > >> > as
> > > > >>> > >> > > > > well.
> > > > >>> > >> > > > > > > > > Perhaps
> > > > >>> > >> > > > > > > > > > > we
> > > > >>> > >> > > > > > > > > > > > can use the
> > UNSUPPORTED_FOR_MESSAGE_FORMAT
> > > > >>> error?
> > > > >>> > >> > > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > > The license thing seems
> straightforward
> > --
> > > > >>> it's
> > > > >>> > >> just a
> > > > >>> > >> > > > matter
> > > > >>> > >> > > > > > of
> > > > >>> > >> > > > > > > > > adding
> > > > >>> > >> > > > > > > > > > > > the text to the right files as per ASF
> > > > >>> guidelines.
> > > > >>> > >> > > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > > With regard to the dictionary support,
> > do
> > > we
> > > > >>> > really
> > > > >>> > >> > need
> > > > >>> > >> > > > that
> > > > >>> > >> > > > > > in
> > > > >>> > >> > > > > > > > the
> > > > >>> > >> > > > > > > > > > > first
> > > > >>> > >> > > > > > > > > > > > version?  Hopefully message batches
> are
> > > big
> > > > >>> enough
> > > > >>> > >> that
> > > > >>> > >> > > > this
> > > > >>> > >> > > > > > > isn't
> > > > >>> > >> > > > > > > > > > > needed.
> > > > >>> > >> > > > > > > > > > > > Some benchmarks might help here.
> > > > >>> > >> > > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > > best,
> > > > >>> > >> > > > > > > > > > > > Colin
> > > > >>> > >> > > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > > On Tue, Aug 7, 2018, at 08:02, Dongjin
> > Lee
> > > > >>> wrote:
> > > > >>> > >> > > > > > > > > > > > > As Kafka 2.0.0 was released, let's
> > > reboot
> > > > >>> this
> > > > >>> > >> issue,
> > > > >>> > >> > > > > KIP-110
> > > > >>> > >> > > > > > > > > > > > > <https://cwiki.apache.org/
> > > > >>> > >> > > confluence/display/KAFKA/KIP-
> > > > >>> > >> > > > > > > > > > > > 110%3A+Add+Codec+for+
> > > ZStandard+Compression>
> > > > >>> > >> > > > > > > > > > > > > .
> > > > >>> > >> > > > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > > > For newcomers, Here is some summary
> of
> > > the
> > > > >>> > >> history:
> > > > >>> > >> > > > KIP-110
> > > > >>> > >> > > > > > was
> > > > >>> > >> > > > > > > > > > > > originally
> > > > >>> > >> > > > > > > > > > > > > worked for the issue KAFKA-4514 but,
> > it
> > > > >>> lacked
> > > > >>> > >> > > benchmark
> > > > >>> > >> > > > > > > results
> > > > >>> > >> > > > > > > > to
> > > > >>> > >> > > > > > > > > > get
> > > > >>> > >> > > > > > > > > > > > the
> > > > >>> > >> > > > > > > > > > > > > agreement of the community. Later,
> > Ivan
> > > > >>> Babrou
> > > > >>> > and
> > > > >>> > >> > some
> > > > >>> > >> > > > > other
> > > > >>> > >> > > > > > > > users
> > > > >>> > >> > > > > > > > > > who
> > > > >>> > >> > > > > > > > > > > > > adopted the patch provided their
> > > excellent
> > > > >>> > >> > performance
> > > > >>> > >> > > > > report
> > > > >>> > >> > > > > > > > which
> > > > >>> > >> > > > > > > > > > is
> > > > >>> > >> > > > > > > > > > > > now
> > > > >>> > >> > > > > > > > > > > > > included in the KIP, but it
> postponed
> > > > again
> > > > >>> > >> because
> > > > >>> > >> > of
> > > > >>> > >> > > > the
> > > > >>> > >> > > > > > > > > community
> > > > >>> > >> > > > > > > > > > > was
> > > > >>> > >> > > > > > > > > > > > > busy for 2.0.0 release. It is why I
> > now
> > > > >>> reboot
> > > > >>> > >> this
> > > > >>> > >> > > > issue.
> > > > >>> > >> > > > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > > > The following is the current status
> of
> > > the
> > > > >>> > >> feature:
> > > > >>> > >> > You
> > > > >>> > >> > > > can
> > > > >>> > >> > > > > > > check
> > > > >>> > >> > > > > > > > > the
> > > > >>> > >> > > > > > > > > > > > > current draft implementation here
> > > > >>> > >> > > > > > > > > > > > > <
> > > > https://github.com/apache/kafka/pull/2267>.
> > > > >>> It
> > > > >>> > >> is
> > > > >>> > >> > > based
> > > > >>> > >> > > > > on
> > > > >>> > >> > > > > > > zstd
> > > > >>> > >> > > > > > > > > > 1.3.5
> > > > >>> > >> > > > > > > > > > > > and
> > > > >>> > >> > > > > > > > > > > > > periodically rebased onto the latest
> > > > >>> trunk[^1].
> > > > >>> > >> > > > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > > > The issues that should be addressed
> is
> > > > like
> > > > >>> > >> > following:
> > > > >>> > >> > > > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > > > *1. Backward Compatibility*
> > > > >>> > >> > > > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > > > To support old consumers, we need to
> > > take
> > > > a
> > > > >>> > >> strategy
> > > > >>> > >> > to
> > > > >>> > >> > > > > > handle
> > > > >>> > >> > > > > > > > the
> > > > >>> > >> > > > > > > > > > old
> > > > >>> > >> > > > > > > > > > > > > consumers. Current candidates are:
> > > > >>> > >> > > > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > > > - Bump API version
> > > > >>> > >> > > > > > > > > > > > > - Leave unchanged: let the old
> clients
> > > > fail.
> > > > >>> > >> > > > > > > > > > > > > - Improve the error messages:
> > > > >>> > >> > > > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > > > *2. Dictionary Support*
> > > > >>> > >> > > > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > > > To support zstd's dictionary feature
> > in
> > > > the
> > > > >>> > future
> > > > >>> > >> > (if
> > > > >>> > >> > > > > > needed),
> > > > >>> > >> > > > > > > > we
> > > > >>> > >> > > > > > > > > > need
> > > > >>> > >> > > > > > > > > > > > to
> > > > >>> > >> > > > > > > > > > > > > sketch how it should be and leave
> some
> > > > room
> > > > >>> for
> > > > >>> > >> it.
> > > > >>> > >> > As
> > > > >>> > >> > > of
> > > > >>> > >> > > > > > now,
> > > > >>> > >> > > > > > > > > there
> > > > >>> > >> > > > > > > > > > > has
> > > > >>> > >> > > > > > > > > > > > > been no discussion on this topic
> yet.
> > > > >>> > >> > > > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > > > *3. License*
> > > > >>> > >> > > > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > > > To use this feature, we need to add
> > > > license
> > > > >>> of
> > > > >>> > >> zstd
> > > > >>> > >> > and
> > > > >>> > >> > > > > > > zstd-jni
> > > > >>> > >> > > > > > > > to
> > > > >>> > >> > > > > > > > > > the
> > > > >>> > >> > > > > > > > > > > > > project. (Thanks to Viktor Somogyi
> for
> > > > >>> raising
> > > > >>> > >> this
> > > > >>> > >> > > > issue!)
> > > > >>> > >> > > > > > It
> > > > >>> > >> > > > > > > > > seems
> > > > >>> > >> > > > > > > > > > > like
> > > > >>> > >> > > > > > > > > > > > > what Apache Spark did would be a
> good
> > > > >>> example
> > > > >>> > but
> > > > >>> > >> > there
> > > > >>> > >> > > > has
> > > > >>> > >> > > > > > > been
> > > > >>> > >> > > > > > > > no
> > > > >>> > >> > > > > > > > > > > > > discussion yet.
> > > > >>> > >> > > > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > > > You can find the details of the
> above
> > > > >>> issues in
> > > > >>> > >> the
> > > > >>> > >> > KIP
> > > > >>> > >> > > > > > > document.
> > > > >>> > >> > > > > > > > > > > Please
> > > > >>> > >> > > > > > > > > > > > > have a look when you are free, and
> > give
> > > me
> > > > >>> > >> feedback.
> > > > >>> > >> > > All
> > > > >>> > >> > > > > > kinds
> > > > >>> > >> > > > > > > of
> > > > >>> > >> > > > > > > > > > > > > participating are welcome.
> > > > >>> > >> > > > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > > > Best,
> > > > >>> > >> > > > > > > > > > > > > Dongjin
> > > > >>> > >> > > > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > > > [^1]: At the time of writing, commit
> > > > >>> 6b4fb8152.
> > > > >>> > >> > > > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > > > On Sat, Jul 14, 2018 at 10:45 PM
> > Dongjin
> > > > >>> Lee <
> > > > >>> > >> > > > > > > dongjin@apache.org
> > > > >>> > >> > > > > > > > >
> > > > >>> > >> > > > > > > > > > > wrote:
> > > > >>> > >> > > > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > > > > Sorry for the late reply.
> > > > >>> > >> > > > > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > > > > In short, I could not submit the
> > > updated
> > > > >>> KIP
> > > > >>> > by
> > > > >>> > >> the
> > > > >>> > >> > > > > feature
> > > > >>> > >> > > > > > > > > freeze
> > > > >>> > >> > > > > > > > > > > > > > deadline of 2.0.0. For this
> reason,
> > it
> > > > >>> will
> > > > >>> > not
> > > > >>> > >> be
> > > > >>> > >> > > > > included
> > > > >>> > >> > > > > > > in
> > > > >>> > >> > > > > > > > > the
> > > > >>> > >> > > > > > > > > > > > 2.0.0
> > > > >>> > >> > > > > > > > > > > > > > release and all discussion for
> this
> > > > issue
> > > > >>> were
> > > > >>> > >> > > > postponed
> > > > >>> > >> > > > > > > after
> > > > >>> > >> > > > > > > > > the
> > > > >>> > >> > > > > > > > > > > > release
> > > > >>> > >> > > > > > > > > > > > > > of 2.0.0.
> > > > >>> > >> > > > > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > > > > I have been updating the PR
> > following
> > > > >>> recent
> > > > >>> > >> > updates.
> > > > >>> > >> > > > > Just
> > > > >>> > >> > > > > > > > now, I
> > > > >>> > >> > > > > > > > > > > > rebased
> > > > >>> > >> > > > > > > > > > > > > > it against the latest trunk and
> > > updated
> > > > >>> the
> > > > >>> > zstd
> > > > >>> > >> > > > version
> > > > >>> > >> > > > > > into
> > > > >>> > >> > > > > > > > > > 1.3.5.
> > > > >>> > >> > > > > > > > > > > > If you
> > > > >>> > >> > > > > > > > > > > > > > need some request, don't hesitate
> to
> > > > >>> notify
> > > > >>> > me.
> > > > >>> > >> > (But
> > > > >>> > >> > > > not
> > > > >>> > >> > > > > > this
> > > > >>> > >> > > > > > > > > > thread
> > > > >>> > >> > > > > > > > > > > -
> > > > >>> > >> > > > > > > > > > > > just
> > > > >>> > >> > > > > > > > > > > > > > send me the message directly.)
> > > > >>> > >> > > > > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > > > > Best,
> > > > >>> > >> > > > > > > > > > > > > > Dongjin
> > > > >>> > >> > > > > > > > > > > > > >
> > > > >>> > >> > > > > > > > > > > > > > On Tue, Jul 10, 2018 at 11:57 PM
> > Bobby
> > > > >>> Evans <
> > > > >>> > >> > > > > > > bobby@apache.org
> > > > >>> > >> > > > > > > > >
> > > > >>> > >> > > > > > > > > > > wrote:
> > > > >>> > >> > > > > > > > > > > > > >
> > > > >>> > >> >

Re: [DISCUSS] KIP-110: Add Codec for ZStandard Compression (Updated)

[Dongjin Lee <do...@apache.org>]

Hi Jason,

You are right. Explicit statements are always better. I updated the
document following your suggestion.

@Magnus

Thanks for the inspection. It seems like a new error code is not a problem.

Thanks,
Dongjin

On Fri, Sep 7, 2018 at 3:23 AM Jason Gustafson <ja...@confluent.io> wrote:

> Hi Dongjin,
>
> The KIP looks good to me. I'd suggest starting a vote. A couple minor
> points that might be worth calling out explicitly in the compatibility
> section:
>
> 1. Zstd will only be allowed for the bumped produce API. For older
> versions, we return UNSUPPORTED_COMPRESSION_TYPE regardless of the message
> format.
> 2. Down-conversion of zstd-compressed records will not be supported.
> Instead we will return UNSUPPORTED_COMPRESSION_TYPE.
>
> Does that sound right?
>
> Thanks,
> Jason
>
> On Thu, Sep 6, 2018 at 1:45 AM, Magnus Edenhill <ma...@edenhill.se>
> wrote:
>
> > > Ismael wrote:
> > > Jason, that's an interesting point regarding the Java client. Do we
> know
> > > what clients in other languages do in these cases?
> >
> > librdkafka (and its bindings) passes unknown/future errors through to the
> > application, the error code remains intact while
> > the error string will be set to something like "Err-123?", which isn't
> very
> > helpful to the user but it at least
> > preserves the original error code for further troubleshooting.
> > For the producer any unknown error returned in the ProduceResponse will
> be
> > considered a permanent delivery failure (no retries),
> > and for the consumer any unknown FetchResponse errors will propagate
> > directly to the application, trigger a fetch backoff, and then
> > continue fetching past that offset.
> >
> > So, from the client's perspective it is not really a problem if new error
> > codes are added to older API versions.
> >
> > /Magnus
> >
> >
> > Den tors 6 sep. 2018 kl 09:45 skrev Dongjin Lee <do...@apache.org>:
> >
> > > I updated the KIP page
> > > <
> > > https://cwiki.apache.org/confluence/display/KAFKA/KIP-
> > 110%3A+Add+Codec+for+ZStandard+Compression
> > > >
> > > following the discussion here. Please take a look when you are free.
> > > If you have any opinion, don't hesitate to give me a message.
> > >
> > > Best,
> > > Dongjin
> > >
> > > On Fri, Aug 31, 2018 at 11:35 PM Dongjin Lee <do...@apache.org>
> wrote:
> > >
> > > > I just updated the draft implementation[^1], rebasing against the
> > latest
> > > > trunk and implementing error routine (i.e., Error code 74 for
> > > > UnsupportedCompressionTypeException.) Since we decided to disallow
> all
> > > > fetch request below version 2.1.0 for the topics specifying
> ZStandard,
> > I
> > > > added an error logic only.
> > > >
> > > > Please have a look when you are free.
> > > >
> > > > Thanks,
> > > > Dongjin
> > > >
> > > > [^1]: Please check the last commit here:
> > > > https://github.com/apache/kafka/pull/2267
> > > >
> > > > On Thu, Aug 23, 2018, 8:55 AM Dongjin Lee <do...@apache.org>
> wrote:
> > > >
> > > >> Jason,
> > > >>
> > > >> Great. +1 for UNSUPPORTED_COMPRESSION_TYPE.
> > > >>
> > > >> Best,
> > > >> Dongjin
> > > >>
> > > >> On Thu, Aug 23, 2018 at 8:19 AM Jason Gustafson <jason@confluent.io
> >
> > > >> wrote:
> > > >>
> > > >>> Hey Dongjin,
> > > >>>
> > > >>> Yeah that's right. For what it's worth, librdkafka also appears to
> > > handle
> > > >>> unexpected error codes. I expect that most client implementations
> > would
> > > >>> either pass through the raw type or convert to an enum using
> > something
> > > >>> like
> > > >>> what the java client does. Since we're expecting the client to fail
> > > >>> anyway,
> > > >>> I'm probably in favor of using the UNSUPPORTED_COMPRESSION_TYPE
> error
> > > >>> code.
> > > >>>
> > > >>> -Jason
> > > >>>
> > > >>> On Wed, Aug 22, 2018 at 1:46 AM, Dongjin Lee <do...@apache.org>
> > > wrote:
> > > >>>
> > > >>> > Jason and Ismael,
> > > >>> >
> > > >>> > It seems like the only thing we need to regard if we define a new
> > > error
> > > >>> > code (i.e., UNSUPPORTED_COMPRESSION_TYPE) would be the
> > implementation
> > > >>> of
> > > >>> > the other language clients, right? At least, this strategy causes
> > any
> > > >>> > problem for Java client. Do I understand correctly?
> > > >>> >
> > > >>> > Thanks,
> > > >>> > Dongjin
> > > >>> >
> > > >>> > On Wed, Aug 22, 2018 at 5:43 PM Dongjin Lee <do...@apache.org>
> > > >>> wrote:
> > > >>> >
> > > >>> > > Jason,
> > > >>> > >
> > > >>> > > > I think we would only use this error code when we /know/ that
> > > zstd
> > > >>> was
> > > >>> > > in use and the client doesn't support it? This is true if
> either
> > 1)
> > > >>> the
> > > >>> > > message needs down-conversion and we encounter a zstd
> compressed
> > > >>> message,
> > > >>> > > or 2) if the topic is explicitly configured to use zstd.
> > > >>> > >
> > > >>> > > Yes, it is right. And you know, the case 1 includes 1.a) old
> > > clients'
> > > >>> > > request v0, v1 records or 1.b) implicit zstd, the compression
> > type
> > > of
> > > >>> > > "producer" with Zstd compressed data.
> > > >>> > >
> > > >>> > > > However, if the compression type is set to "producer," then
> the
> > > >>> fetched
> > > >>> > > data may or may not be compressed with zstd. In this case, we
> > > return
> > > >>> the
> > > >>> > > data to the client and expect it to fail parsing. Is that
> > correct?
> > > >>> > >
> > > >>> > > Exactly.
> > > >>> > >
> > > >>> > > Following your message, I reviewed the implementation of
> > > >>> > > `KafkaApis#handleFetchRequest,` which handles the fetch
> request.
> > > And
> > > >>> > found
> > > >>> > > that the information we can use is like the following:
> > > >>> > >
> > > >>> > > 1. Client's fetch request version. (`versionId` variable)
> > > >>> > > 2. Log's compression type. (`logConfig` variable)
> > > >>> > >
> > > >>> > > We can't detect the actual compression type of the data, unless
> > we
> > > >>> > inspect
> > > >>> > > the `RecordBatch` included in the `Records` instance (i.e.,
> > > >>> > > `unconvertedRecords` variable.) Since it requires some
> > performance
> > > >>> issue,
> > > >>> > > it is not our option - in short, we can't be sure if given
> chunks
> > > of
> > > >>> data
> > > >>> > > are compressed with zstd or not.
> > > >>> > >
> > > >>> > > So, conclusion: we can return an error in the case of 1.a and 2
> > > >>> easily,
> > > >>> > > with the information above. In the case 1.b (implicit zstd), we
> > can
> > > >>> just
> > > >>> > > return the data by do nothing special and expect it to fail
> > > parsing.
> > > >>> > >
> > > >>> > > Thanks,
> > > >>> > > Dongjin
> > > >>> > >
> > > >>> > > On Wed, Aug 22, 2018 at 12:02 PM Ismael Juma <
> ismaelj@gmail.com>
> > > >>> wrote:
> > > >>> > >
> > > >>> > >> Jason, that's an interesting point regarding the Java client.
> Do
> > > we
> > > >>> know
> > > >>> > >> what clients in other languages do in these cases?
> > > >>> > >>
> > > >>> > >> Ismael
> > > >>> > >>
> > > >>> > >> On Tue, 21 Aug 2018, 17:30 Jason Gustafson, <
> jason@confluent.io
> > >
> > > >>> wrote:
> > > >>> > >>
> > > >>> > >> > Hi Dongjin,
> > > >>> > >> >
> > > >>> > >> > One of the complications is that old versions of the API
> will
> > > not
> > > >>> > >> expect a
> > > >>> > >> > new error code. However, since we expect this to be a fatal
> > > error
> > > >>> > anyway
> > > >>> > >> > for old clients, it may still be more useful to return the
> > > correct
> > > >>> > error
> > > >>> > >> > code. For example, the Kafka clients use the following code
> to
> > > >>> convert
> > > >>> > >> the
> > > >>> > >> > error code:
> > > >>> > >> >
> > > >>> > >> >     public static Errors forCode(short code) {
> > > >>> > >> >         Errors error = codeToError.get(code);
> > > >>> > >> >         if (error != null) {
> > > >>> > >> >             return error;
> > > >>> > >> >         } else {
> > > >>> > >> >             log.warn("Unexpected error code: {}.", code);
> > > >>> > >> >             return UNKNOWN_SERVER_ERROR;
> > > >>> > >> >         }
> > > >>> > >> >     }
> > > >>> > >> >
> > > >>> > >> > If we return an unsupported error code, it will be converted
> > to
> > > an
> > > >>> > >> UNKNOWN
> > > >>> > >> > error, but at least we will get the message in the log with
> > the
> > > >>> > correct
> > > >>> > >> > code. That seems preferable to returning a misleading error
> > > code.
> > > >>> So I
> > > >>> > >> > wonder if we can use the new UNSUPPORTED_COMPRESSION_TYPE
> > error
> > > >>> even
> > > >>> > for
> > > >>> > >> > older versions.
> > > >>> > >> >
> > > >>> > >> > Also, one question just to check my understanding. I think
> we
> > > >>> would
> > > >>> > only
> > > >>> > >> > use this error code when we /know/ that zstd was in use and
> > the
> > > >>> client
> > > >>> > >> > doesn't support it? This is true if either 1) the message
> > needs
> > > >>> > >> > down-conversion and we encounter a zstd compressed message,
> or
> > > 2)
> > > >>> if
> > > >>> > the
> > > >>> > >> > topic is explicitly configured to use zstd. However, if the
> > > >>> > compression
> > > >>> > >> > type is set to "producer," then the fetched data may or may
> > not
> > > be
> > > >>> > >> > compressed with zstd. In this case, we return the data to
> the
> > > >>> client
> > > >>> > and
> > > >>> > >> > expect it to fail parsing. Is that correct?
> > > >>> > >> >
> > > >>> > >> > Thanks,
> > > >>> > >> > Jason
> > > >>> > >> >
> > > >>> > >> >
> > > >>> > >> >
> > > >>> > >> > On Tue, Aug 21, 2018 at 9:08 AM, Dongjin Lee <
> > > dongjin@apache.org>
> > > >>> > >> wrote:
> > > >>> > >> >
> > > >>> > >> > > Ismael, Jason and all,
> > > >>> > >> > >
> > > >>> > >> > > I rewrote the backward compatibility strategy & its
> > > alternatives
> > > >>> > like
> > > >>> > >> > > following, based on Ismael & Jason's comments. Since it is
> > not
> > > >>> > >> updated to
> > > >>> > >> > > the wiki yet, don't hesitate to give me a message if you
> > have
> > > >>> any
> > > >>> > >> opinion
> > > >>> > >> > > on it.
> > > >>> > >> > >
> > > >>> > >> > > ```
> > > >>> > >> > > *Backward Compatibility*
> > > >>> > >> > >
> > > >>> > >> > > We need to establish some backward-compatibility strategy
> > for
> > > >>> the
> > > >>> > >> case an
> > > >>> > >> > > old client subscribes a topic using ZStandard implicitly
> > > (i.e.,
> > > >>> > >> > > 'compression.type' configuration of given topic is
> > 'producer'
> > > >>> and
> > > >>> > the
> > > >>> > >> > > producer compressed the records with ZStandard). We have
> the
> > > >>> > following
> > > >>> > >> > > options for this situation:
> > > >>> > >> > >
> > > >>> > >> > > *A. Support ZStandard to the old clients which can
> > understand
> > > >>> v0, v1
> > > >>> > >> > > messages only.*
> > > >>> > >> > >
> > > >>> > >> > > This strategy necessarily requires the down-conversion of
> v2
> > > >>> message
> > > >>> > >> > > compressed with Zstandard into v0 or v1 messages, which
> > means
> > > a
> > > >>> > >> > > considerable performance degradation. So we rejected this
> > > >>> strategy.
> > > >>> > >> > >
> > > >>> > >> > > *B. Bump the API version and support only v2-available
> > > clients*
> > > >>> > >> > >
> > > >>> > >> > > With this approach, we can message the old clients that
> they
> > > >>> are old
> > > >>> > >> and
> > > >>> > >> > > should be upgraded. However, there are still several
> options
> > > >>> for the
> > > >>> > >> > Error
> > > >>> > >> > > code.
> > > >>> > >> > >
> > > >>> > >> > > *B.1. INVALID_REQUEST (42)*
> > > >>> > >> > >
> > > >>> > >> > > This option gives the client so little information; the
> user
> > > >>> can be
> > > >>> > >> > > confused about why the client worked correctly in the past
> > > >>> suddenly
> > > >>> > >> > > encounters a problem. So we rejected this strategy.
> > > >>> > >> > >
> > > >>> > >> > > *B.2. CORRUPT_MESSAGE (2)*
> > > >>> > >> > >
> > > >>> > >> > > This option gives inaccurate information; the user can be
> > > >>> surprised
> > > >>> > >> and
> > > >>> > >> > > misunderstand that the log files are broken in some way.
> So
> > we
> > > >>> > >> rejected
> > > >>> > >> > > this strategy.
> > > >>> > >> > >
> > > >>> > >> > > *B.3 UNSUPPORTED_FOR_MESSAGE_FORMAT (43)*
> > > >>> > >> > >
> > > >>> > >> > > The advantage of this approach is we don't need to define
> a
> > > new
> > > >>> > error
> > > >>> > >> > code;
> > > >>> > >> > > we can reuse it and that's all.
> > > >>> > >> > >
> > > >>> > >> > > The disadvantage of this approach is that it is also a
> > little
> > > >>> bit
> > > >>> > >> vague;
> > > >>> > >> > > This error code is defined as a work for KIP-98[^1] and
> now
> > > >>> returned
> > > >>> > >> in
> > > >>> > >> > the
> > > >>> > >> > > transaction error.
> > > >>> > >> > >
> > > >>> > >> > > *B.4. UNSUPPORTED_COMPRESSION_TYPE (new)*
> > > >>> > >> > >
> > > >>> > >> > > The advantage of this approach is that it is clear and
> > > provides
> > > >>> an
> > > >>> > >> exact
> > > >>> > >> > > description. The disadvantage is we need to add a new
> error
> > > >>> code.
> > > >>> > >> > > ```
> > > >>> > >> > >
> > > >>> > >> > > *It seems like what we need to choose is now so clear:
> > > >>> > >> > > UNSUPPORTED_FOR_MESSAGE_FORMAT (B.3) or
> > > >>> UNSUPPORTED_COMPRESSION_TYPE
> > > >>> > >> > > (B.4).*
> > > >>> > >> > > The first one doesn't need a new error message but the
> > latter
> > > is
> > > >>> > more
> > > >>> > >> > > explicit. Which one do you prefer? Since all of you have
> > much
> > > >>> more
> > > >>> > >> > > experience and knowledge than me, I will follow your
> > decision.
> > > >>> The
> > > >>> > >> wiki
> > > >>> > >> > > page will be updated following the decision also.
> > > >>> > >> > >
> > > >>> > >> > > Best,
> > > >>> > >> > > Dongjin
> > > >>> > >> > >
> > > >>> > >> > > [^1]: https://issues.apache.org/jira/browse/KAFKA-4990
> > > >>> > >> > >
> > > >>> > >> > > On Sun, Aug 19, 2018 at 4:58 AM Ismael Juma <
> > > ismaelj@gmail.com>
> > > >>> > >> wrote:
> > > >>> > >> > >
> > > >>> > >> > > > Sounds reasonable to me.
> > > >>> > >> > > >
> > > >>> > >> > > > Ismael
> > > >>> > >> > > >
> > > >>> > >> > > > On Sat, 18 Aug 2018, 12:20 Jason Gustafson, <
> > > >>> jason@confluent.io>
> > > >>> > >> > wrote:
> > > >>> > >> > > >
> > > >>> > >> > > > > Hey Ismael,
> > > >>> > >> > > > >
> > > >>> > >> > > > > Your summary looks good to me. I think it might also
> be
> > a
> > > >>> good
> > > >>> > >> idea
> > > >>> > >> > to
> > > >>> > >> > > > add
> > > >>> > >> > > > > a new UNSUPPORTED_COMPRESSION_TYPE error code to go
> > along
> > > >>> with
> > > >>> > the
> > > >>> > >> > > > version
> > > >>> > >> > > > > bumps. We won't be able to use it for old api versions
> > > >>> since the
> > > >>> > >> > > clients
> > > >>> > >> > > > > will not understand it, but we can use it going
> forward
> > so
> > > >>> that
> > > >>> > >> we're
> > > >>> > >> > > not
> > > >>> > >> > > > > stuck in a similar situation with a new message format
> > and
> > > >>> a new
> > > >>> > >> > codec
> > > >>> > >> > > to
> > > >>> > >> > > > > support. Another option is to use
> > UNSUPPORTED_FOR_MESSAGE_
> > > >>> > FORMAT,
> > > >>> > >> but
> > > >>> > >> > > it
> > > >>> > >> > > > is
> > > >>> > >> > > > > not as explicit.
> > > >>> > >> > > > >
> > > >>> > >> > > > > -Jason
> > > >>> > >> > > > >
> > > >>> > >> > > > > On Fri, Aug 17, 2018 at 5:19 PM, Ismael Juma <
> > > >>> ismael@juma.me.uk
> > > >>> > >
> > > >>> > >> > > wrote:
> > > >>> > >> > > > >
> > > >>> > >> > > > > > Hi Dongjin and Jason,
> > > >>> > >> > > > > >
> > > >>> > >> > > > > > I would agree. My summary:
> > > >>> > >> > > > > >
> > > >>> > >> > > > > > 1. Support zstd with message format 2 only.
> > > >>> > >> > > > > > 2. Bump produce and fetch request versions.
> > > >>> > >> > > > > > 3. Provide broker errors whenever possible based on
> > the
> > > >>> > request
> > > >>> > >> > > version
> > > >>> > >> > > > > and
> > > >>> > >> > > > > > rely on clients for the cases where the broker can't
> > > >>> validate
> > > >>> > >> > > > efficiently
> > > >>> > >> > > > > > (example message format 2 consumer that supports the
> > > >>> latest
> > > >>> > >> fetch
> > > >>> > >> > > > version
> > > >>> > >> > > > > > but doesn't support zstd).
> > > >>> > >> > > > > >
> > > >>> > >> > > > > > If there's general agreement on this, I suggest we
> > > update
> > > >>> the
> > > >>> > >> KIP
> > > >>> > >> > to
> > > >>> > >> > > > > state
> > > >>> > >> > > > > > the proposal and to move the rejected options to its
> > own
> > > >>> > >> section.
> > > >>> > >> > And
> > > >>> > >> > > > > then
> > > >>> > >> > > > > > start a vote!
> > > >>> > >> > > > > >
> > > >>> > >> > > > > > Ismael
> > > >>> > >> > > > > >
> > > >>> > >> > > > > > On Fri, Aug 17, 2018 at 4:00 PM Jason Gustafson <
> > > >>> > >> > jason@confluent.io>
> > > >>> > >> > > > > > wrote:
> > > >>> > >> > > > > >
> > > >>> > >> > > > > > > Hi Dongjin,
> > > >>> > >> > > > > > >
> > > >>> > >> > > > > > > Yes, that's a good summary. For clients which
> > support
> > > >>> v2,
> > > >>> > the
> > > >>> > >> > > client
> > > >>> > >> > > > > can
> > > >>> > >> > > > > > > parse the message format and hopefully raise a
> > useful
> > > >>> error
> > > >>> > >> > message
> > > >>> > >> > > > > > > indicating the unsupported compression type. For
> > older
> > > >>> > >> clients,
> > > >>> > >> > our
> > > >>> > >> > > > > > options
> > > >>> > >> > > > > > > are probably (1) to down-convert to the old format
> > > >>> using no
> > > >>> > >> > > > compression
> > > >>> > >> > > > > > > type, or (2) to return an error code. I'm leaning
> > > >>> toward the
> > > >>> > >> > latter
> > > >>> > >> > > > as
> > > >>> > >> > > > > > the
> > > >>> > >> > > > > > > simpler solution, but the challenge is finding a
> > good
> > > >>> error
> > > >>> > >> code.
> > > >>> > >> > > Two
> > > >>> > >> > > > > > > possibilities might be INVALID_REQUEST or
> > > >>> CORRUPT_MESSAGE.
> > > >>> > The
> > > >>> > >> > > > downside
> > > >>> > >> > > > > > is
> > > >>> > >> > > > > > > that old clients probably won't get a helpful
> > message.
> > > >>> > >> However,
> > > >>> > >> > at
> > > >>> > >> > > > > least
> > > >>> > >> > > > > > > the behavior will be consistent in the sense that
> > all
> > > >>> > clients
> > > >>> > >> > will
> > > >>> > >> > > > fail
> > > >>> > >> > > > > > if
> > > >>> > >> > > > > > > they do not support zstandard.
> > > >>> > >> > > > > > >
> > > >>> > >> > > > > > > What do you think?
> > > >>> > >> > > > > > >
> > > >>> > >> > > > > > > Thanks,
> > > >>> > >> > > > > > > Jason
> > > >>> > >> > > > > > >
> > > >>> > >> > > > > > > On Fri, Aug 17, 2018 at 8:08 AM, Dongjin Lee <
> > > >>> > >> dongjin@apache.org
> > > >>> > >> > >
> > > >>> > >> > > > > wrote:
> > > >>> > >> > > > > > >
> > > >>> > >> > > > > > > > Thanks Jason, I reviewed the down-converting
> logic
> > > >>> > following
> > > >>> > >> > your
> > > >>> > >> > > > > > > > explanation.[^1] You mean the following
> routines,
> > > >>> right?
> > > >>> > >> > > > > > > >
> > > >>> > >> > > > > > > > -
> > > >>> > >> > > > > > > >
> > > https://github.com/apache/kafka/blob/trunk/core/src/
> > > >>> > >> > > > > > > > main/scala/kafka/server/KafkaApis.scala#L534
> > > >>> > >> > > > > > > > -
> > > >>> > >> > > > > > > >
> > > >>> https://github.com/apache/kafka/blob/trunk/clients/src/
> > > >>> > >> > > > > > > > main/java/org/apache/kafka/common/record/
> > > >>> > >> > > LazyDownConversionRecords.
> > > >>> > >> > > > > > > > java#L165
> > > >>> > >> > > > > > > > -
> > > >>> > >> > > > > > > >
> > > >>> https://github.com/apache/kafka/blob/trunk/clients/src/
> > > >>> > >> > > > > > > >
> > > >>> > >> main/java/org/apache/kafka/common/record/RecordsUtil.java#L40
> > > >>> > >> > > > > > > >
> > > >>> > >> > > > > > > > It seems like your stance is like following:
> > > >>> > >> > > > > > > >
> > > >>> > >> > > > > > > > 1. In principle, Kafka does not change the
> > > compression
> > > >>> > codec
> > > >>> > >> > when
> > > >>> > >> > > > > > > > down-converting, since it requires inspecting
> the
> > > >>> fetched
> > > >>> > >> data,
> > > >>> > >> > > > which
> > > >>> > >> > > > > > is
> > > >>> > >> > > > > > > > expensive.
> > > >>> > >> > > > > > > > 2. However, there are some cases the fetched
> data
> > is
> > > >>> > >> inspected
> > > >>> > >> > > > > anyway.
> > > >>> > >> > > > > > In
> > > >>> > >> > > > > > > > this case, we can provide compression conversion
> > > from
> > > >>> > >> Zstandard
> > > >>> > >> > > to
> > > >>> > >> > > > > > > > classical ones[^2].
> > > >>> > >> > > > > > > >
> > > >>> > >> > > > > > > > And from what I understand, the cases where the
> > > client
> > > >>> > >> without
> > > >>> > >> > > > > > ZStandard
> > > >>> > >> > > > > > > > support receives ZStandard compressed records
> can
> > be
> > > >>> > >> organized
> > > >>> > >> > > into
> > > >>> > >> > > > > two
> > > >>> > >> > > > > > > > cases:
> > > >>> > >> > > > > > > >
> > > >>> > >> > > > > > > > a. The 'compression.type' configuration of given
> > > >>> topic is
> > > >>> > >> > > > 'producer'
> > > >>> > >> > > > > > and
> > > >>> > >> > > > > > > > the producer compressed the records with
> > ZStandard.
> > > >>> (that
> > > >>> > >> is,
> > > >>> > >> > > using
> > > >>> > >> > > > > > > > ZStandard implicitly.)
> > > >>> > >> > > > > > > > b.  The 'compression.type' configuration of
> given
> > > >>> topic is
> > > >>> > >> > > 'zstd';
> > > >>> > >> > > > > that
> > > >>> > >> > > > > > > is,
> > > >>> > >> > > > > > > > using ZStandard explicitly.
> > > >>> > >> > > > > > > >
> > > >>> > >> > > > > > > > As you stated, we don't have to handle the case
> b
> > > >>> > specially.
> > > >>> > >> > So,
> > > >>> > >> > > It
> > > >>> > >> > > > > > seems
> > > >>> > >> > > > > > > > like we can narrow the focus of the problem by
> > > joining
> > > >>> > case
> > > >>> > >> 1
> > > >>> > >> > and
> > > >>> > >> > > > > case
> > > >>> > >> > > > > > b
> > > >>> > >> > > > > > > > like the following:
> > > >>> > >> > > > > > > >
> > > >>> > >> > > > > > > > > Given the topic with 'producer' as its
> > > >>> > 'compression.type'
> > > >>> > >> > > > > > > configuration,
> > > >>> > >> > > > > > > > ZStandard compressed records and old client
> > without
> > > >>> > >> ZStandard,
> > > >>> > >> > is
> > > >>> > >> > > > > there
> > > >>> > >> > > > > > > any
> > > >>> > >> > > > > > > > case we need to inspect the records and can
> change
> > > the
> > > >>> > >> > > compression
> > > >>> > >> > > > > > type?
> > > >>> > >> > > > > > > If
> > > >>> > >> > > > > > > > so, can we provide compression type converting?
> > > >>> > >> > > > > > > >
> > > >>> > >> > > > > > > > Do I understand correctly?
> > > >>> > >> > > > > > > >
> > > >>> > >> > > > > > > > Best,
> > > >>> > >> > > > > > > > Dongjin
> > > >>> > >> > > > > > > >
> > > >>> > >> > > > > > > > [^1]: I'm sorry, I found that I was a little bit
> > > >>> > >> > misunderstanding
> > > >>> > >> > > > how
> > > >>> > >> > > > > > API
> > > >>> > >> > > > > > > > version works, after reviewing the downconvert
> > logic
> > > >>> & the
> > > >>> > >> > > protocol
> > > >>> > >> > > > > > > > documentation <
> https://kafka.apache.org/protocol
> > >.
> > > >>> > >> > > > > > > > [^2]: None, Gzip, Snappy, Lz4.
> > > >>> > >> > > > > > > >
> > > >>> > >> > > > > > > > On Tue, Aug 14, 2018 at 2:16 AM Jason Gustafson
> <
> > > >>> > >> > > > jason@confluent.io>
> > > >>> > >> > > > > > > > wrote:
> > > >>> > >> > > > > > > >
> > > >>> > >> > > > > > > > > >
> > > >>> > >> > > > > > > > > > But in my opinion, since the client will
> fail
> > > >>> with the
> > > >>> > >> API
> > > >>> > >> > > > > version,
> > > >>> > >> > > > > > > so
> > > >>> > >> > > > > > > > we
> > > >>> > >> > > > > > > > > > don't need to down-convert the messages
> > anyway.
> > > >>> Isn't
> > > >>> > >> it?
> > > >>> > >> > > So, I
> > > >>> > >> > > > > > think
> > > >>> > >> > > > > > > > we
> > > >>> > >> > > > > > > > > > don't care about this case. (I'm sorry, I am
> > not
> > > >>> > >> familiar
> > > >>> > >> > > with
> > > >>> > >> > > > > > > > > down-convert
> > > >>> > >> > > > > > > > > > logic.)
> > > >>> > >> > > > > > > > >
> > > >>> > >> > > > > > > > >
> > > >>> > >> > > > > > > > > Currently the broker down-converts
> automatically
> > > >>> when it
> > > >>> > >> > > receives
> > > >>> > >> > > > > an
> > > >>> > >> > > > > > > old
> > > >>> > >> > > > > > > > > version of the fetch request (a version which
> is
> > > >>> known
> > > >>> > to
> > > >>> > >> > > predate
> > > >>> > >> > > > > the
> > > >>> > >> > > > > > > > > message format in use). Typically when
> > > >>> down-converting
> > > >>> > the
> > > >>> > >> > > > message
> > > >>> > >> > > > > > > > format,
> > > >>> > >> > > > > > > > > we use the same compression type, but there is
> > not
> > > >>> much
> > > >>> > >> point
> > > >>> > >> > > in
> > > >>> > >> > > > > > doing
> > > >>> > >> > > > > > > so
> > > >>> > >> > > > > > > > > when we know the client doesn't support it. So
> > if
> > > >>> > >> zstandard
> > > >>> > >> > is
> > > >>> > >> > > in
> > > >>> > >> > > > > > use,
> > > >>> > >> > > > > > > > and
> > > >>> > >> > > > > > > > > we have to down-convert anyway, then we can
> > choose
> > > >>> to
> > > >>> > use
> > > >>> > >> a
> > > >>> > >> > > > > different
> > > >>> > >> > > > > > > > > compression type or no compression type.
> > > >>> > >> > > > > > > > >
> > > >>> > >> > > > > > > > > From my perspective, there is no significant
> > > >>> downside to
> > > >>> > >> > > bumping
> > > >>> > >> > > > > the
> > > >>> > >> > > > > > > > > protocol version and it has several potential
> > > >>> benefits.
> > > >>> > >> > Version
> > > >>> > >> > > > > bumps
> > > >>> > >> > > > > > > are
> > > >>> > >> > > > > > > > > cheap. The main question mark in my mind is
> > about
> > > >>> > >> > > > down-conversion.
> > > >>> > >> > > > > > > > Figuring
> > > >>> > >> > > > > > > > > out whether down-conversion is needed is hard
> > > >>> generally
> > > >>> > >> > without
> > > >>> > >> > > > > > > > inspecting
> > > >>> > >> > > > > > > > > the fetched data, which is expensive. I think
> we
> > > >>> agree
> > > >>> > in
> > > >>> > >> > > > principle
> > > >>> > >> > > > > > > that
> > > >>> > >> > > > > > > > we
> > > >>> > >> > > > > > > > > do not want to have to pay this cost generally
> > and
> > > >>> > prefer
> > > >>> > >> the
> > > >>> > >> > > > > clients
> > > >>> > >> > > > > > > to
> > > >>> > >> > > > > > > > > fail when they see an unhandled compression
> > type.
> > > >>> The
> > > >>> > >> point I
> > > >>> > >> > > was
> > > >>> > >> > > > > > > making
> > > >>> > >> > > > > > > > is
> > > >>> > >> > > > > > > > > that there are some cases where we are either
> > > >>> inspecting
> > > >>> > >> the
> > > >>> > >> > > data
> > > >>> > >> > > > > > > anyway
> > > >>> > >> > > > > > > > > (because we have to down-convert the message
> > > >>> format), or
> > > >>> > >> we
> > > >>> > >> > > have
> > > >>> > >> > > > an
> > > >>> > >> > > > > > > easy
> > > >>> > >> > > > > > > > > way to tell whether zstandard is in use (the
> > topic
> > > >>> has
> > > >>> > it
> > > >>> > >> > > > > configured
> > > >>> > >> > > > > > > > > explicitly). In the latter case, we don't have
> > to
> > > >>> handle
> > > >>> > >> it
> > > >>> > >> > > > > > specially.
> > > >>> > >> > > > > > > > But
> > > >>> > >> > > > > > > > > we do have to decide how we will handle
> > > >>> down-conversion
> > > >>> > to
> > > >>> > >> > > older
> > > >>> > >> > > > > > > formats.
> > > >>> > >> > > > > > > > >
> > > >>> > >> > > > > > > > > -Jason
> > > >>> > >> > > > > > > > >
> > > >>> > >> > > > > > > > > On Sun, Aug 12, 2018 at 5:15 PM, Dongjin Lee <
> > > >>> > >> > > dongjin@apache.org
> > > >>> > >> > > > >
> > > >>> > >> > > > > > > wrote:
> > > >>> > >> > > > > > > > >
> > > >>> > >> > > > > > > > > > Colin and Jason,
> > > >>> > >> > > > > > > > > >
> > > >>> > >> > > > > > > > > > Thanks for your opinions. In summarizing,
> the
> > > >>> Pros and
> > > >>> > >> Cons
> > > >>> > >> > > of
> > > >>> > >> > > > > > > bumping
> > > >>> > >> > > > > > > > > > fetch API version are:
> > > >>> > >> > > > > > > > > >
> > > >>> > >> > > > > > > > > > Cons:
> > > >>> > >> > > > > > > > > >
> > > >>> > >> > > > > > > > > > - The Broker can't know whether a given
> > message
> > > >>> batch
> > > >>> > is
> > > >>> > >> > > > > compressed
> > > >>> > >> > > > > > > > with
> > > >>> > >> > > > > > > > > > zstd or not.
> > > >>> > >> > > > > > > > > > - Need some additional logic for the topic
> > > >>> explicitly
> > > >>> > >> > > > configured
> > > >>> > >> > > > > to
> > > >>> > >> > > > > > > use
> > > >>> > >> > > > > > > > > > zstd.
> > > >>> > >> > > > > > > > > >
> > > >>> > >> > > > > > > > > > Pros:
> > > >>> > >> > > > > > > > > >
> > > >>> > >> > > > > > > > > > - The broker doesn't need to conduct
> expensive
> > > >>> > >> > > down-conversion.
> > > >>> > >> > > > > > > > > > - Can message the users to update their
> > client.
> > > >>> > >> > > > > > > > > >
> > > >>> > >> > > > > > > > > > So, opinions for the backward-compatibility
> > > >>> policy by
> > > >>> > >> far:
> > > >>> > >> > > > > > > > > >
> > > >>> > >> > > > > > > > > > - A: bump the API version - +2 (Colin,
> Jason)
> > > >>> > >> > > > > > > > > > - B: leave unchanged - +1 (Viktor)
> > > >>> > >> > > > > > > > > >
> > > >>> > >> > > > > > > > > > Here are my additional comments:
> > > >>> > >> > > > > > > > > >
> > > >>> > >> > > > > > > > > > @Colin
> > > >>> > >> > > > > > > > > >
> > > >>> > >> > > > > > > > > > I greatly appreciate your response. In the
> > case
> > > >>> of the
> > > >>> > >> > > > dictionary
> > > >>> > >> > > > > > > > > support,
> > > >>> > >> > > > > > > > > > of course, this issue should be addressed
> > later
> > > >>> so we
> > > >>> > >> don't
> > > >>> > >> > > > need
> > > >>> > >> > > > > it
> > > >>> > >> > > > > > > in
> > > >>> > >> > > > > > > > > the
> > > >>> > >> > > > > > > > > > first version. You are right - it is not
> late
> > to
> > > >>> try
> > > >>> > it
> > > >>> > >> > after
> > > >>> > >> > > > > some
> > > >>> > >> > > > > > > > > > benchmarks. What I mean is, we should keep
> in
> > > >>> mind on
> > > >>> > >> that
> > > >>> > >> > > > > > potential
> > > >>> > >> > > > > > > > > > feature.
> > > >>> > >> > > > > > > > > >
> > > >>> > >> > > > > > > > > > @Jason
> > > >>> > >> > > > > > > > > >
> > > >>> > >> > > > > > > > > > You wrote,
> > > >>> > >> > > > > > > > > >
> > > >>> > >> > > > > > > > > > > Similarly, if we have to down-convert
> anyway
> > > >>> because
> > > >>> > >> the
> > > >>> > >> > > > client
> > > >>> > >> > > > > > > does
> > > >>> > >> > > > > > > > > not
> > > >>> > >> > > > > > > > > > understand the message format, then we could
> > > also
> > > >>> use
> > > >>> > a
> > > >>> > >> > > > different
> > > >>> > >> > > > > > > > > > compression type.
> > > >>> > >> > > > > > > > > >
> > > >>> > >> > > > > > > > > > But in my opinion, since the client will
> fail
> > > >>> with the
> > > >>> > >> API
> > > >>> > >> > > > > version,
> > > >>> > >> > > > > > > so
> > > >>> > >> > > > > > > > we
> > > >>> > >> > > > > > > > > > don't need to down-convert the messages
> > anyway.
> > > >>> Isn't
> > > >>> > >> it?
> > > >>> > >> > > So, I
> > > >>> > >> > > > > > think
> > > >>> > >> > > > > > > > we
> > > >>> > >> > > > > > > > > > don't care about this case. (I'm sorry, I am
> > not
> > > >>> > >> familiar
> > > >>> > >> > > with
> > > >>> > >> > > > > > > > > down-convert
> > > >>> > >> > > > > > > > > > logic.)
> > > >>> > >> > > > > > > > > >
> > > >>> > >> > > > > > > > > > Please give more opinions. Thanks!
> > > >>> > >> > > > > > > > > >
> > > >>> > >> > > > > > > > > > - Dongjin
> > > >>> > >> > > > > > > > > >
> > > >>> > >> > > > > > > > > >
> > > >>> > >> > > > > > > > > > On Wed, Aug 8, 2018 at 6:41 AM Jason
> > Gustafson <
> > > >>> > >> > > > > jason@confluent.io
> > > >>> > >> > > > > > >
> > > >>> > >> > > > > > > > > wrote:
> > > >>> > >> > > > > > > > > >
> > > >>> > >> > > > > > > > > > > Hey Colin,
> > > >>> > >> > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > The problem for the fetch API is that the
> > > broker
> > > >>> > does
> > > >>> > >> not
> > > >>> > >> > > > > > generally
> > > >>> > >> > > > > > > > > know
> > > >>> > >> > > > > > > > > > if
> > > >>> > >> > > > > > > > > > > a batch was compressed with zstd unless it
> > > >>> parses
> > > >>> > it.
> > > >>> > >> I
> > > >>> > >> > > think
> > > >>> > >> > > > > the
> > > >>> > >> > > > > > > > goal
> > > >>> > >> > > > > > > > > > here
> > > >>> > >> > > > > > > > > > > is to avoid the expensive down-conversion
> > that
> > > >>> is
> > > >>> > >> needed
> > > >>> > >> > to
> > > >>> > >> > > > > > ensure
> > > >>> > >> > > > > > > > > > > compatibility because it is only necessary
> > if
> > > >>> zstd
> > > >>> > is
> > > >>> > >> > > > actually
> > > >>> > >> > > > > in
> > > >>> > >> > > > > > > > use.
> > > >>> > >> > > > > > > > > > But
> > > >>> > >> > > > > > > > > > > as long as old clients can parse the
> message
> > > >>> format,
> > > >>> > >> they
> > > >>> > >> > > > > should
> > > >>> > >> > > > > > > get
> > > >>> > >> > > > > > > > a
> > > >>> > >> > > > > > > > > > > reasonable error if they see an
> unsupported
> > > >>> > >> compression
> > > >>> > >> > > type
> > > >>> > >> > > > in
> > > >>> > >> > > > > > the
> > > >>> > >> > > > > > > > > > > attributes. Basically the onus is on users
> > to
> > > >>> ensure
> > > >>> > >> that
> > > >>> > >> > > > their
> > > >>> > >> > > > > > > > > consumers
> > > >>> > >> > > > > > > > > > > have been updated prior to using zstd. It
> > > seems
> > > >>> > like a
> > > >>> > >> > > > > reasonable
> > > >>> > >> > > > > > > > > > tradeoff
> > > >>> > >> > > > > > > > > > > to me. There are a couple cases that might
> > be
> > > >>> worth
> > > >>> > >> > > thinking
> > > >>> > >> > > > > > > through:
> > > >>> > >> > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > 1. If a topic is explicitly configured to
> > use
> > > >>> zstd,
> > > >>> > >> then
> > > >>> > >> > we
> > > >>> > >> > > > > don't
> > > >>> > >> > > > > > > > need
> > > >>> > >> > > > > > > > > to
> > > >>> > >> > > > > > > > > > > check the fetched data for the compression
> > > type
> > > >>> to
> > > >>> > >> know
> > > >>> > >> > if
> > > >>> > >> > > we
> > > >>> > >> > > > > > need
> > > >>> > >> > > > > > > > > > > down-conversion. If we did bump the Fetch
> > API
> > > >>> > version,
> > > >>> > >> > then
> > > >>> > >> > > > we
> > > >>> > >> > > > > > > could
> > > >>> > >> > > > > > > > > > handle
> > > >>> > >> > > > > > > > > > > this case by either down-converting using
> a
> > > >>> > different
> > > >>> > >> > > > > compression
> > > >>> > >> > > > > > > > type
> > > >>> > >> > > > > > > > > or
> > > >>> > >> > > > > > > > > > > returning an error.
> > > >>> > >> > > > > > > > > > > 2. Similarly, if we have to down-convert
> > > anyway
> > > >>> > >> because
> > > >>> > >> > the
> > > >>> > >> > > > > > client
> > > >>> > >> > > > > > > > does
> > > >>> > >> > > > > > > > > > not
> > > >>> > >> > > > > > > > > > > understand the message format, then we
> could
> > > >>> also
> > > >>> > use
> > > >>> > >> a
> > > >>> > >> > > > > different
> > > >>> > >> > > > > > > > > > > compression type.
> > > >>> > >> > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > For the produce API, I think it's
> reasonable
> > > to
> > > >>> bump
> > > >>> > >> the
> > > >>> > >> > > api
> > > >>> > >> > > > > > > version.
> > > >>> > >> > > > > > > > > > This
> > > >>> > >> > > > > > > > > > > can be used by clients to check whether a
> > > broker
> > > >>> > >> supports
> > > >>> > >> > > > zstd.
> > > >>> > >> > > > > > For
> > > >>> > >> > > > > > > > > > > example, we might support a list of
> > preferred
> > > >>> > >> compression
> > > >>> > >> > > > types
> > > >>> > >> > > > > > in
> > > >>> > >> > > > > > > > the
> > > >>> > >> > > > > > > > > > > producer and we could use the broker to
> > detect
> > > >>> which
> > > >>> > >> > > version
> > > >>> > >> > > > to
> > > >>> > >> > > > > > > use.
> > > >>> > >> > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > -Jason
> > > >>> > >> > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > On Tue, Aug 7, 2018 at 1:32 PM, Colin
> > McCabe <
> > > >>> > >> > > > > cmccabe@apache.org
> > > >>> > >> > > > > > >
> > > >>> > >> > > > > > > > > wrote:
> > > >>> > >> > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > > Thanks for bumping this, Dongjin.  ZStd
> > is a
> > > >>> good
> > > >>> > >> > > > compression
> > > >>> > >> > > > > > > codec
> > > >>> > >> > > > > > > > > > and I
> > > >>> > >> > > > > > > > > > > > hope we can get this support in soon!
> > > >>> > >> > > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > > I would say we can just bump the API
> > version
> > > >>> to
> > > >>> > >> > indicate
> > > >>> > >> > > > that
> > > >>> > >> > > > > > > ZStd
> > > >>> > >> > > > > > > > > > > support
> > > >>> > >> > > > > > > > > > > > is expected in new clients.  We probably
> > > need
> > > >>> some
> > > >>> > >> way
> > > >>> > >> > of
> > > >>> > >> > > > > > > > indicating
> > > >>> > >> > > > > > > > > to
> > > >>> > >> > > > > > > > > > > the
> > > >>> > >> > > > > > > > > > > > older clients that they can't consume
> the
> > > >>> > >> partitions,
> > > >>> > >> > as
> > > >>> > >> > > > > well.
> > > >>> > >> > > > > > > > > Perhaps
> > > >>> > >> > > > > > > > > > > we
> > > >>> > >> > > > > > > > > > > > can use the
> UNSUPPORTED_FOR_MESSAGE_FORMAT
> > > >>> error?
> > > >>> > >> > > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > > The license thing seems straightforward
> --
> > > >>> it's
> > > >>> > >> just a
> > > >>> > >> > > > matter
> > > >>> > >> > > > > > of
> > > >>> > >> > > > > > > > > adding
> > > >>> > >> > > > > > > > > > > > the text to the right files as per ASF
> > > >>> guidelines.
> > > >>> > >> > > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > > With regard to the dictionary support,
> do
> > we
> > > >>> > really
> > > >>> > >> > need
> > > >>> > >> > > > that
> > > >>> > >> > > > > > in
> > > >>> > >> > > > > > > > the
> > > >>> > >> > > > > > > > > > > first
> > > >>> > >> > > > > > > > > > > > version?  Hopefully message batches are
> > big
> > > >>> enough
> > > >>> > >> that
> > > >>> > >> > > > this
> > > >>> > >> > > > > > > isn't
> > > >>> > >> > > > > > > > > > > needed.
> > > >>> > >> > > > > > > > > > > > Some benchmarks might help here.
> > > >>> > >> > > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > > best,
> > > >>> > >> > > > > > > > > > > > Colin
> > > >>> > >> > > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > > On Tue, Aug 7, 2018, at 08:02, Dongjin
> Lee
> > > >>> wrote:
> > > >>> > >> > > > > > > > > > > > > As Kafka 2.0.0 was released, let's
> > reboot
> > > >>> this
> > > >>> > >> issue,
> > > >>> > >> > > > > KIP-110
> > > >>> > >> > > > > > > > > > > > > <https://cwiki.apache.org/
> > > >>> > >> > > confluence/display/KAFKA/KIP-
> > > >>> > >> > > > > > > > > > > > 110%3A+Add+Codec+for+
> > ZStandard+Compression>
> > > >>> > >> > > > > > > > > > > > > .
> > > >>> > >> > > > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > > > For newcomers, Here is some summary of
> > the
> > > >>> > >> history:
> > > >>> > >> > > > KIP-110
> > > >>> > >> > > > > > was
> > > >>> > >> > > > > > > > > > > > originally
> > > >>> > >> > > > > > > > > > > > > worked for the issue KAFKA-4514 but,
> it
> > > >>> lacked
> > > >>> > >> > > benchmark
> > > >>> > >> > > > > > > results
> > > >>> > >> > > > > > > > to
> > > >>> > >> > > > > > > > > > get
> > > >>> > >> > > > > > > > > > > > the
> > > >>> > >> > > > > > > > > > > > > agreement of the community. Later,
> Ivan
> > > >>> Babrou
> > > >>> > and
> > > >>> > >> > some
> > > >>> > >> > > > > other
> > > >>> > >> > > > > > > > users
> > > >>> > >> > > > > > > > > > who
> > > >>> > >> > > > > > > > > > > > > adopted the patch provided their
> > excellent
> > > >>> > >> > performance
> > > >>> > >> > > > > report
> > > >>> > >> > > > > > > > which
> > > >>> > >> > > > > > > > > > is
> > > >>> > >> > > > > > > > > > > > now
> > > >>> > >> > > > > > > > > > > > > included in the KIP, but it postponed
> > > again
> > > >>> > >> because
> > > >>> > >> > of
> > > >>> > >> > > > the
> > > >>> > >> > > > > > > > > community
> > > >>> > >> > > > > > > > > > > was
> > > >>> > >> > > > > > > > > > > > > busy for 2.0.0 release. It is why I
> now
> > > >>> reboot
> > > >>> > >> this
> > > >>> > >> > > > issue.
> > > >>> > >> > > > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > > > The following is the current status of
> > the
> > > >>> > >> feature:
> > > >>> > >> > You
> > > >>> > >> > > > can
> > > >>> > >> > > > > > > check
> > > >>> > >> > > > > > > > > the
> > > >>> > >> > > > > > > > > > > > > current draft implementation here
> > > >>> > >> > > > > > > > > > > > > <
> > > https://github.com/apache/kafka/pull/2267>.
> > > >>> It
> > > >>> > >> is
> > > >>> > >> > > based
> > > >>> > >> > > > > on
> > > >>> > >> > > > > > > zstd
> > > >>> > >> > > > > > > > > > 1.3.5
> > > >>> > >> > > > > > > > > > > > and
> > > >>> > >> > > > > > > > > > > > > periodically rebased onto the latest
> > > >>> trunk[^1].
> > > >>> > >> > > > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > > > The issues that should be addressed is
> > > like
> > > >>> > >> > following:
> > > >>> > >> > > > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > > > *1. Backward Compatibility*
> > > >>> > >> > > > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > > > To support old consumers, we need to
> > take
> > > a
> > > >>> > >> strategy
> > > >>> > >> > to
> > > >>> > >> > > > > > handle
> > > >>> > >> > > > > > > > the
> > > >>> > >> > > > > > > > > > old
> > > >>> > >> > > > > > > > > > > > > consumers. Current candidates are:
> > > >>> > >> > > > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > > > - Bump API version
> > > >>> > >> > > > > > > > > > > > > - Leave unchanged: let the old clients
> > > fail.
> > > >>> > >> > > > > > > > > > > > > - Improve the error messages:
> > > >>> > >> > > > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > > > *2. Dictionary Support*
> > > >>> > >> > > > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > > > To support zstd's dictionary feature
> in
> > > the
> > > >>> > future
> > > >>> > >> > (if
> > > >>> > >> > > > > > needed),
> > > >>> > >> > > > > > > > we
> > > >>> > >> > > > > > > > > > need
> > > >>> > >> > > > > > > > > > > > to
> > > >>> > >> > > > > > > > > > > > > sketch how it should be and leave some
> > > room
> > > >>> for
> > > >>> > >> it.
> > > >>> > >> > As
> > > >>> > >> > > of
> > > >>> > >> > > > > > now,
> > > >>> > >> > > > > > > > > there
> > > >>> > >> > > > > > > > > > > has
> > > >>> > >> > > > > > > > > > > > > been no discussion on this topic yet.
> > > >>> > >> > > > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > > > *3. License*
> > > >>> > >> > > > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > > > To use this feature, we need to add
> > > license
> > > >>> of
> > > >>> > >> zstd
> > > >>> > >> > and
> > > >>> > >> > > > > > > zstd-jni
> > > >>> > >> > > > > > > > to
> > > >>> > >> > > > > > > > > > the
> > > >>> > >> > > > > > > > > > > > > project. (Thanks to Viktor Somogyi for
> > > >>> raising
> > > >>> > >> this
> > > >>> > >> > > > issue!)
> > > >>> > >> > > > > > It
> > > >>> > >> > > > > > > > > seems
> > > >>> > >> > > > > > > > > > > like
> > > >>> > >> > > > > > > > > > > > > what Apache Spark did would be a good
> > > >>> example
> > > >>> > but
> > > >>> > >> > there
> > > >>> > >> > > > has
> > > >>> > >> > > > > > > been
> > > >>> > >> > > > > > > > no
> > > >>> > >> > > > > > > > > > > > > discussion yet.
> > > >>> > >> > > > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > > > You can find the details of the above
> > > >>> issues in
> > > >>> > >> the
> > > >>> > >> > KIP
> > > >>> > >> > > > > > > document.
> > > >>> > >> > > > > > > > > > > Please
> > > >>> > >> > > > > > > > > > > > > have a look when you are free, and
> give
> > me
> > > >>> > >> feedback.
> > > >>> > >> > > All
> > > >>> > >> > > > > > kinds
> > > >>> > >> > > > > > > of
> > > >>> > >> > > > > > > > > > > > > participating are welcome.
> > > >>> > >> > > > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > > > Best,
> > > >>> > >> > > > > > > > > > > > > Dongjin
> > > >>> > >> > > > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > > > [^1]: At the time of writing, commit
> > > >>> 6b4fb8152.
> > > >>> > >> > > > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > > > On Sat, Jul 14, 2018 at 10:45 PM
> Dongjin
> > > >>> Lee <
> > > >>> > >> > > > > > > dongjin@apache.org
> > > >>> > >> > > > > > > > >
> > > >>> > >> > > > > > > > > > > wrote:
> > > >>> > >> > > > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > > > > Sorry for the late reply.
> > > >>> > >> > > > > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > > > > In short, I could not submit the
> > updated
> > > >>> KIP
> > > >>> > by
> > > >>> > >> the
> > > >>> > >> > > > > feature
> > > >>> > >> > > > > > > > > freeze
> > > >>> > >> > > > > > > > > > > > > > deadline of 2.0.0. For this reason,
> it
> > > >>> will
> > > >>> > not
> > > >>> > >> be
> > > >>> > >> > > > > included
> > > >>> > >> > > > > > > in
> > > >>> > >> > > > > > > > > the
> > > >>> > >> > > > > > > > > > > > 2.0.0
> > > >>> > >> > > > > > > > > > > > > > release and all discussion for this
> > > issue
> > > >>> were
> > > >>> > >> > > > postponed
> > > >>> > >> > > > > > > after
> > > >>> > >> > > > > > > > > the
> > > >>> > >> > > > > > > > > > > > release
> > > >>> > >> > > > > > > > > > > > > > of 2.0.0.
> > > >>> > >> > > > > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > > > > I have been updating the PR
> following
> > > >>> recent
> > > >>> > >> > updates.
> > > >>> > >> > > > > Just
> > > >>> > >> > > > > > > > now, I
> > > >>> > >> > > > > > > > > > > > rebased
> > > >>> > >> > > > > > > > > > > > > > it against the latest trunk and
> > updated
> > > >>> the
> > > >>> > zstd
> > > >>> > >> > > > version
> > > >>> > >> > > > > > into
> > > >>> > >> > > > > > > > > > 1.3.5.
> > > >>> > >> > > > > > > > > > > > If you
> > > >>> > >> > > > > > > > > > > > > > need some request, don't hesitate to
> > > >>> notify
> > > >>> > me.
> > > >>> > >> > (But
> > > >>> > >> > > > not
> > > >>> > >> > > > > > this
> > > >>> > >> > > > > > > > > > thread
> > > >>> > >> > > > > > > > > > > -
> > > >>> > >> > > > > > > > > > > > just
> > > >>> > >> > > > > > > > > > > > > > send me the message directly.)
> > > >>> > >> > > > > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > > > > Best,
> > > >>> > >> > > > > > > > > > > > > > Dongjin
> > > >>> > >> > > > > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > > > > On Tue, Jul 10, 2018 at 11:57 PM
> Bobby
> > > >>> Evans <
> > > >>> > >> > > > > > > bobby@apache.org
> > > >>> > >> > > > > > > > >
> > > >>> > >> > > > > > > > > > > wrote:
> > > >>> > >> > > > > > > > > > > > > >
> > > >>> > >> > > > > > > > > > > > > >> I there any update on this.  The
> > > >>> performance
> > > >>> > >> > > > > improvements
> > > >>> > >> > > > > > > are
> > > >>> > >> > > > > > > > > > quite
> > > >>> > >> > > > > > > > > > > > > >> impressive and I really would like
> to
> > > >>> stop
> > > >>> > >> forking
> > > >>> > >> > > > kafka
> > > >>> > >> > > > > > > just
> > > >>> > >> > > > > > > > to
> > > >>> > >> > > > > > > > > > get
> > > >>> > >> > > > > > > > > > > > this
> > > >>> > >> > > > > > > > > > > > > >> in.
> > > >>> > >> > > > > > > > > > > > > >>
> > > >>> > >> > > > > > > > > > > > > >> Thanks,
> > > >>> > >> > > > > > > > > > > > > >>
> > > >>> > >> > > > > > > > > > > > > >> Bobby
> > > >>> > >> > > > > > > > > > > > > >>
> > > >>> > >> > > > > > > > > > > > > >> On Wed, Jun 13, 2018 at 8:56 PM
> > Dongjin
> > > >>> Lee <
> > > >>> > >> > > > > > > > dongjin@apache.org
> > > >>> > >> > > > > > > > > >
> > > >>> > >> > > > > > > > > > > > wrote:
> > > >>> > >> > > > > > > > > > > > > >>
> > > >>> > >> > > > > > > > > > > > > >> > Ismael,
> > > >>> > >> > > > > > > > > > > > > >> >
> > > >>> > >> > > > > > > > > > > > > >> > Oh, I forgot all of you are on
> > > working
> > > >>> > frenzy
> > > >>> > >> > for
> > > >>> > >> > > > 2.0!
> > > >>> > >> > > > > > No
> > > >>> > >> > > > > > > > > > problem,
> > > >>> > >> > > > > > > > > > > > take
> > > >>> > >> > > > > > > > > > > > > >> > your time. I am also working at
> > > another
> > > >>> > issue
> > > >>> > >> > now.
> > > >>> > >> > > > > Thank
> > > >>> > >> > > > > > > you
> > > >>> > >> > > > > > > > > for
> > > >>> > >> > > > > > > > > > > > > >> letting me
> > > >>> > >> > > > > > > > > > > > > >> > know.
> > > >>> > >> > > > > > > > > > > > > >> >
> > > >>> > >> > > > > > > > > > > > > >> > Best,
> > > >>> > >> > > > > > > > > > > > > >> > Dongjin
> > > >>> > >> > > > > > > > > > > > > >> >
> > > >>> > >> > > > > > > > > > > > > >> > On Wed, Jun 13, 2018, 11:44 PM
> > Ismael
> > > >>> Juma
> > > >>> > <
> > > >>> > >> > > > > > > > ismaelj@gmail.com
> > > >>> > >> > > > > > > > > >
> > > >>> > >> > > > > > > > > > > > wrote:
> > > >>> > >> > > > > > > > > > > > > >> >
> > > >>> > >> > > > > > > > > > > > > >> > > Sorry for the delay Dongjin.
> > > >>> Everyone is
> > > >>> > >> busy
> > > >>> > >> > > > > > finalising
> > > >>> > >> > > > > > > > > > 2.0.0.
> > > >>> > >> > > > > > > > > > > > This
> > > >>> > >> > > > > > > > > > > > > >> KIP
> > > >>> > >> > > > > > > > > > > > > >> > > seems like a great candidate
> for
> > > >>> 2.1.0
> > > >>> > and
> > > >>> > >> > > > hopefully
> > > >>> > >> > > > > > > there
> > > >>> > >> > > > > > > > > > will
> > > >>> > >> > > > > > > > > > > be
> > > >>> > >> > > > > > > > > > > > > >> more
> > > >>> > >> > > > > > > > > > > > > >> > of
> > > >>> > >> > > > > > > > > > > > > >> > > a discussion next week. :)
> > > >>> > >> > > > > > > > > > > > > >> > >
> > > >>> > >> > > > > > > > > > > > > >> > > Ismael
> > > >>> > >> > > > > > > > > > > > > >> > >
> > > >>> > >> > > > > > > > > > > > > >> > > On Wed, 13 Jun 2018, 05:17
> > Dongjin
> > > >>> Lee, <
> > > >>> > >> > > > > > > > dongjin@apache.org
> > > >>> > >> > > > > > > > > >
> > > >>> > >> > > > > > > > > > > > wrote:
> > > >>> > >> > > > > > > > > > > > > >> > >
> > > >>> > >> > > > > > > > > > > > > >> > > > Hello. I just updated my
> draft
>


-- 
*Dongjin Lee*

*A hitchhiker in the mathematical world.*

*github:  <http://goog_969573159/>github.com/dongjinleekr
<http://github.com/dongjinleekr>linkedin: kr.linkedin.com/in/dongjinleekr
<http://kr.linkedin.com/in/dongjinleekr>slideshare:
www.slideshare.net/dongjinleekr
<http://www.slideshare.net/dongjinleekr>*

Re: [DISCUSS] KIP-110: Add Codec for ZStandard Compression (Updated)

[Jason Gustafson <ja...@confluent.io>]

Hi Dongjin,

The KIP looks good to me. I'd suggest starting a vote. A couple minor
points that might be worth calling out explicitly in the compatibility
section:

1. Zstd will only be allowed for the bumped produce API. For older
versions, we return UNSUPPORTED_COMPRESSION_TYPE regardless of the message
format.
2. Down-conversion of zstd-compressed records will not be supported.
Instead we will return UNSUPPORTED_COMPRESSION_TYPE.

Does that sound right?

Thanks,
Jason

On Thu, Sep 6, 2018 at 1:45 AM, Magnus Edenhill <ma...@edenhill.se> wrote:

> > Ismael wrote:
> > Jason, that's an interesting point regarding the Java client. Do we know
> > what clients in other languages do in these cases?
>
> librdkafka (and its bindings) passes unknown/future errors through to the
> application, the error code remains intact while
> the error string will be set to something like "Err-123?", which isn't very
> helpful to the user but it at least
> preserves the original error code for further troubleshooting.
> For the producer any unknown error returned in the ProduceResponse will be
> considered a permanent delivery failure (no retries),
> and for the consumer any unknown FetchResponse errors will propagate
> directly to the application, trigger a fetch backoff, and then
> continue fetching past that offset.
>
> So, from the client's perspective it is not really a problem if new error
> codes are added to older API versions.
>
> /Magnus
>
>
> Den tors 6 sep. 2018 kl 09:45 skrev Dongjin Lee <do...@apache.org>:
>
> > I updated the KIP page
> > <
> > https://cwiki.apache.org/confluence/display/KAFKA/KIP-
> 110%3A+Add+Codec+for+ZStandard+Compression
> > >
> > following the discussion here. Please take a look when you are free.
> > If you have any opinion, don't hesitate to give me a message.
> >
> > Best,
> > Dongjin
> >
> > On Fri, Aug 31, 2018 at 11:35 PM Dongjin Lee <do...@apache.org> wrote:
> >
> > > I just updated the draft implementation[^1], rebasing against the
> latest
> > > trunk and implementing error routine (i.e., Error code 74 for
> > > UnsupportedCompressionTypeException.) Since we decided to disallow all
> > > fetch request below version 2.1.0 for the topics specifying ZStandard,
> I
> > > added an error logic only.
> > >
> > > Please have a look when you are free.
> > >
> > > Thanks,
> > > Dongjin
> > >
> > > [^1]: Please check the last commit here:
> > > https://github.com/apache/kafka/pull/2267
> > >
> > > On Thu, Aug 23, 2018, 8:55 AM Dongjin Lee <do...@apache.org> wrote:
> > >
> > >> Jason,
> > >>
> > >> Great. +1 for UNSUPPORTED_COMPRESSION_TYPE.
> > >>
> > >> Best,
> > >> Dongjin
> > >>
> > >> On Thu, Aug 23, 2018 at 8:19 AM Jason Gustafson <ja...@confluent.io>
> > >> wrote:
> > >>
> > >>> Hey Dongjin,
> > >>>
> > >>> Yeah that's right. For what it's worth, librdkafka also appears to
> > handle
> > >>> unexpected error codes. I expect that most client implementations
> would
> > >>> either pass through the raw type or convert to an enum using
> something
> > >>> like
> > >>> what the java client does. Since we're expecting the client to fail
> > >>> anyway,
> > >>> I'm probably in favor of using the UNSUPPORTED_COMPRESSION_TYPE error
> > >>> code.
> > >>>
> > >>> -Jason
> > >>>
> > >>> On Wed, Aug 22, 2018 at 1:46 AM, Dongjin Lee <do...@apache.org>
> > wrote:
> > >>>
> > >>> > Jason and Ismael,
> > >>> >
> > >>> > It seems like the only thing we need to regard if we define a new
> > error
> > >>> > code (i.e., UNSUPPORTED_COMPRESSION_TYPE) would be the
> implementation
> > >>> of
> > >>> > the other language clients, right? At least, this strategy causes
> any
> > >>> > problem for Java client. Do I understand correctly?
> > >>> >
> > >>> > Thanks,
> > >>> > Dongjin
> > >>> >
> > >>> > On Wed, Aug 22, 2018 at 5:43 PM Dongjin Lee <do...@apache.org>
> > >>> wrote:
> > >>> >
> > >>> > > Jason,
> > >>> > >
> > >>> > > > I think we would only use this error code when we /know/ that
> > zstd
> > >>> was
> > >>> > > in use and the client doesn't support it? This is true if either
> 1)
> > >>> the
> > >>> > > message needs down-conversion and we encounter a zstd compressed
> > >>> message,
> > >>> > > or 2) if the topic is explicitly configured to use zstd.
> > >>> > >
> > >>> > > Yes, it is right. And you know, the case 1 includes 1.a) old
> > clients'
> > >>> > > request v0, v1 records or 1.b) implicit zstd, the compression
> type
> > of
> > >>> > > "producer" with Zstd compressed data.
> > >>> > >
> > >>> > > > However, if the compression type is set to "producer," then the
> > >>> fetched
> > >>> > > data may or may not be compressed with zstd. In this case, we
> > return
> > >>> the
> > >>> > > data to the client and expect it to fail parsing. Is that
> correct?
> > >>> > >
> > >>> > > Exactly.
> > >>> > >
> > >>> > > Following your message, I reviewed the implementation of
> > >>> > > `KafkaApis#handleFetchRequest,` which handles the fetch request.
> > And
> > >>> > found
> > >>> > > that the information we can use is like the following:
> > >>> > >
> > >>> > > 1. Client's fetch request version. (`versionId` variable)
> > >>> > > 2. Log's compression type. (`logConfig` variable)
> > >>> > >
> > >>> > > We can't detect the actual compression type of the data, unless
> we
> > >>> > inspect
> > >>> > > the `RecordBatch` included in the `Records` instance (i.e.,
> > >>> > > `unconvertedRecords` variable.) Since it requires some
> performance
> > >>> issue,
> > >>> > > it is not our option - in short, we can't be sure if given chunks
> > of
> > >>> data
> > >>> > > are compressed with zstd or not.
> > >>> > >
> > >>> > > So, conclusion: we can return an error in the case of 1.a and 2
> > >>> easily,
> > >>> > > with the information above. In the case 1.b (implicit zstd), we
> can
> > >>> just
> > >>> > > return the data by do nothing special and expect it to fail
> > parsing.
> > >>> > >
> > >>> > > Thanks,
> > >>> > > Dongjin
> > >>> > >
> > >>> > > On Wed, Aug 22, 2018 at 12:02 PM Ismael Juma <is...@gmail.com>
> > >>> wrote:
> > >>> > >
> > >>> > >> Jason, that's an interesting point regarding the Java client. Do
> > we
> > >>> know
> > >>> > >> what clients in other languages do in these cases?
> > >>> > >>
> > >>> > >> Ismael
> > >>> > >>
> > >>> > >> On Tue, 21 Aug 2018, 17:30 Jason Gustafson, <jason@confluent.io
> >
> > >>> wrote:
> > >>> > >>
> > >>> > >> > Hi Dongjin,
> > >>> > >> >
> > >>> > >> > One of the complications is that old versions of the API will
> > not
> > >>> > >> expect a
> > >>> > >> > new error code. However, since we expect this to be a fatal
> > error
> > >>> > anyway
> > >>> > >> > for old clients, it may still be more useful to return the
> > correct
> > >>> > error
> > >>> > >> > code. For example, the Kafka clients use the following code to
> > >>> convert
> > >>> > >> the
> > >>> > >> > error code:
> > >>> > >> >
> > >>> > >> >     public static Errors forCode(short code) {
> > >>> > >> >         Errors error = codeToError.get(code);
> > >>> > >> >         if (error != null) {
> > >>> > >> >             return error;
> > >>> > >> >         } else {
> > >>> > >> >             log.warn("Unexpected error code: {}.", code);
> > >>> > >> >             return UNKNOWN_SERVER_ERROR;
> > >>> > >> >         }
> > >>> > >> >     }
> > >>> > >> >
> > >>> > >> > If we return an unsupported error code, it will be converted
> to
> > an
> > >>> > >> UNKNOWN
> > >>> > >> > error, but at least we will get the message in the log with
> the
> > >>> > correct
> > >>> > >> > code. That seems preferable to returning a misleading error
> > code.
> > >>> So I
> > >>> > >> > wonder if we can use the new UNSUPPORTED_COMPRESSION_TYPE
> error
> > >>> even
> > >>> > for
> > >>> > >> > older versions.
> > >>> > >> >
> > >>> > >> > Also, one question just to check my understanding. I think we
> > >>> would
> > >>> > only
> > >>> > >> > use this error code when we /know/ that zstd was in use and
> the
> > >>> client
> > >>> > >> > doesn't support it? This is true if either 1) the message
> needs
> > >>> > >> > down-conversion and we encounter a zstd compressed message, or
> > 2)
> > >>> if
> > >>> > the
> > >>> > >> > topic is explicitly configured to use zstd. However, if the
> > >>> > compression
> > >>> > >> > type is set to "producer," then the fetched data may or may
> not
> > be
> > >>> > >> > compressed with zstd. In this case, we return the data to the
> > >>> client
> > >>> > and
> > >>> > >> > expect it to fail parsing. Is that correct?
> > >>> > >> >
> > >>> > >> > Thanks,
> > >>> > >> > Jason
> > >>> > >> >
> > >>> > >> >
> > >>> > >> >
> > >>> > >> > On Tue, Aug 21, 2018 at 9:08 AM, Dongjin Lee <
> > dongjin@apache.org>
> > >>> > >> wrote:
> > >>> > >> >
> > >>> > >> > > Ismael, Jason and all,
> > >>> > >> > >
> > >>> > >> > > I rewrote the backward compatibility strategy & its
> > alternatives
> > >>> > like
> > >>> > >> > > following, based on Ismael & Jason's comments. Since it is
> not
> > >>> > >> updated to
> > >>> > >> > > the wiki yet, don't hesitate to give me a message if you
> have
> > >>> any
> > >>> > >> opinion
> > >>> > >> > > on it.
> > >>> > >> > >
> > >>> > >> > > ```
> > >>> > >> > > *Backward Compatibility*
> > >>> > >> > >
> > >>> > >> > > We need to establish some backward-compatibility strategy
> for
> > >>> the
> > >>> > >> case an
> > >>> > >> > > old client subscribes a topic using ZStandard implicitly
> > (i.e.,
> > >>> > >> > > 'compression.type' configuration of given topic is
> 'producer'
> > >>> and
> > >>> > the
> > >>> > >> > > producer compressed the records with ZStandard). We have the
> > >>> > following
> > >>> > >> > > options for this situation:
> > >>> > >> > >
> > >>> > >> > > *A. Support ZStandard to the old clients which can
> understand
> > >>> v0, v1
> > >>> > >> > > messages only.*
> > >>> > >> > >
> > >>> > >> > > This strategy necessarily requires the down-conversion of v2
> > >>> message
> > >>> > >> > > compressed with Zstandard into v0 or v1 messages, which
> means
> > a
> > >>> > >> > > considerable performance degradation. So we rejected this
> > >>> strategy.
> > >>> > >> > >
> > >>> > >> > > *B. Bump the API version and support only v2-available
> > clients*
> > >>> > >> > >
> > >>> > >> > > With this approach, we can message the old clients that they
> > >>> are old
> > >>> > >> and
> > >>> > >> > > should be upgraded. However, there are still several options
> > >>> for the
> > >>> > >> > Error
> > >>> > >> > > code.
> > >>> > >> > >
> > >>> > >> > > *B.1. INVALID_REQUEST (42)*
> > >>> > >> > >
> > >>> > >> > > This option gives the client so little information; the user
> > >>> can be
> > >>> > >> > > confused about why the client worked correctly in the past
> > >>> suddenly
> > >>> > >> > > encounters a problem. So we rejected this strategy.
> > >>> > >> > >
> > >>> > >> > > *B.2. CORRUPT_MESSAGE (2)*
> > >>> > >> > >
> > >>> > >> > > This option gives inaccurate information; the user can be
> > >>> surprised
> > >>> > >> and
> > >>> > >> > > misunderstand that the log files are broken in some way. So
> we
> > >>> > >> rejected
> > >>> > >> > > this strategy.
> > >>> > >> > >
> > >>> > >> > > *B.3 UNSUPPORTED_FOR_MESSAGE_FORMAT (43)*
> > >>> > >> > >
> > >>> > >> > > The advantage of this approach is we don't need to define a
> > new
> > >>> > error
> > >>> > >> > code;
> > >>> > >> > > we can reuse it and that's all.
> > >>> > >> > >
> > >>> > >> > > The disadvantage of this approach is that it is also a
> little
> > >>> bit
> > >>> > >> vague;
> > >>> > >> > > This error code is defined as a work for KIP-98[^1] and now
> > >>> returned
> > >>> > >> in
> > >>> > >> > the
> > >>> > >> > > transaction error.
> > >>> > >> > >
> > >>> > >> > > *B.4. UNSUPPORTED_COMPRESSION_TYPE (new)*
> > >>> > >> > >
> > >>> > >> > > The advantage of this approach is that it is clear and
> > provides
> > >>> an
> > >>> > >> exact
> > >>> > >> > > description. The disadvantage is we need to add a new error
> > >>> code.
> > >>> > >> > > ```
> > >>> > >> > >
> > >>> > >> > > *It seems like what we need to choose is now so clear:
> > >>> > >> > > UNSUPPORTED_FOR_MESSAGE_FORMAT (B.3) or
> > >>> UNSUPPORTED_COMPRESSION_TYPE
> > >>> > >> > > (B.4).*
> > >>> > >> > > The first one doesn't need a new error message but the
> latter
> > is
> > >>> > more
> > >>> > >> > > explicit. Which one do you prefer? Since all of you have
> much
> > >>> more
> > >>> > >> > > experience and knowledge than me, I will follow your
> decision.
> > >>> The
> > >>> > >> wiki
> > >>> > >> > > page will be updated following the decision also.
> > >>> > >> > >
> > >>> > >> > > Best,
> > >>> > >> > > Dongjin
> > >>> > >> > >
> > >>> > >> > > [^1]: https://issues.apache.org/jira/browse/KAFKA-4990
> > >>> > >> > >
> > >>> > >> > > On Sun, Aug 19, 2018 at 4:58 AM Ismael Juma <
> > ismaelj@gmail.com>
> > >>> > >> wrote:
> > >>> > >> > >
> > >>> > >> > > > Sounds reasonable to me.
> > >>> > >> > > >
> > >>> > >> > > > Ismael
> > >>> > >> > > >
> > >>> > >> > > > On Sat, 18 Aug 2018, 12:20 Jason Gustafson, <
> > >>> jason@confluent.io>
> > >>> > >> > wrote:
> > >>> > >> > > >
> > >>> > >> > > > > Hey Ismael,
> > >>> > >> > > > >
> > >>> > >> > > > > Your summary looks good to me. I think it might also be
> a
> > >>> good
> > >>> > >> idea
> > >>> > >> > to
> > >>> > >> > > > add
> > >>> > >> > > > > a new UNSUPPORTED_COMPRESSION_TYPE error code to go
> along
> > >>> with
> > >>> > the
> > >>> > >> > > > version
> > >>> > >> > > > > bumps. We won't be able to use it for old api versions
> > >>> since the
> > >>> > >> > > clients
> > >>> > >> > > > > will not understand it, but we can use it going forward
> so
> > >>> that
> > >>> > >> we're
> > >>> > >> > > not
> > >>> > >> > > > > stuck in a similar situation with a new message format
> and
> > >>> a new
> > >>> > >> > codec
> > >>> > >> > > to
> > >>> > >> > > > > support. Another option is to use
> UNSUPPORTED_FOR_MESSAGE_
> > >>> > FORMAT,
> > >>> > >> but
> > >>> > >> > > it
> > >>> > >> > > > is
> > >>> > >> > > > > not as explicit.
> > >>> > >> > > > >
> > >>> > >> > > > > -Jason
> > >>> > >> > > > >
> > >>> > >> > > > > On Fri, Aug 17, 2018 at 5:19 PM, Ismael Juma <
> > >>> ismael@juma.me.uk
> > >>> > >
> > >>> > >> > > wrote:
> > >>> > >> > > > >
> > >>> > >> > > > > > Hi Dongjin and Jason,
> > >>> > >> > > > > >
> > >>> > >> > > > > > I would agree. My summary:
> > >>> > >> > > > > >
> > >>> > >> > > > > > 1. Support zstd with message format 2 only.
> > >>> > >> > > > > > 2. Bump produce and fetch request versions.
> > >>> > >> > > > > > 3. Provide broker errors whenever possible based on
> the
> > >>> > request
> > >>> > >> > > version
> > >>> > >> > > > > and
> > >>> > >> > > > > > rely on clients for the cases where the broker can't
> > >>> validate
> > >>> > >> > > > efficiently
> > >>> > >> > > > > > (example message format 2 consumer that supports the
> > >>> latest
> > >>> > >> fetch
> > >>> > >> > > > version
> > >>> > >> > > > > > but doesn't support zstd).
> > >>> > >> > > > > >
> > >>> > >> > > > > > If there's general agreement on this, I suggest we
> > update
> > >>> the
> > >>> > >> KIP
> > >>> > >> > to
> > >>> > >> > > > > state
> > >>> > >> > > > > > the proposal and to move the rejected options to its
> own
> > >>> > >> section.
> > >>> > >> > And
> > >>> > >> > > > > then
> > >>> > >> > > > > > start a vote!
> > >>> > >> > > > > >
> > >>> > >> > > > > > Ismael
> > >>> > >> > > > > >
> > >>> > >> > > > > > On Fri, Aug 17, 2018 at 4:00 PM Jason Gustafson <
> > >>> > >> > jason@confluent.io>
> > >>> > >> > > > > > wrote:
> > >>> > >> > > > > >
> > >>> > >> > > > > > > Hi Dongjin,
> > >>> > >> > > > > > >
> > >>> > >> > > > > > > Yes, that's a good summary. For clients which
> support
> > >>> v2,
> > >>> > the
> > >>> > >> > > client
> > >>> > >> > > > > can
> > >>> > >> > > > > > > parse the message format and hopefully raise a
> useful
> > >>> error
> > >>> > >> > message
> > >>> > >> > > > > > > indicating the unsupported compression type. For
> older
> > >>> > >> clients,
> > >>> > >> > our
> > >>> > >> > > > > > options
> > >>> > >> > > > > > > are probably (1) to down-convert to the old format
> > >>> using no
> > >>> > >> > > > compression
> > >>> > >> > > > > > > type, or (2) to return an error code. I'm leaning
> > >>> toward the
> > >>> > >> > latter
> > >>> > >> > > > as
> > >>> > >> > > > > > the
> > >>> > >> > > > > > > simpler solution, but the challenge is finding a
> good
> > >>> error
> > >>> > >> code.
> > >>> > >> > > Two
> > >>> > >> > > > > > > possibilities might be INVALID_REQUEST or
> > >>> CORRUPT_MESSAGE.
> > >>> > The
> > >>> > >> > > > downside
> > >>> > >> > > > > > is
> > >>> > >> > > > > > > that old clients probably won't get a helpful
> message.
> > >>> > >> However,
> > >>> > >> > at
> > >>> > >> > > > > least
> > >>> > >> > > > > > > the behavior will be consistent in the sense that
> all
> > >>> > clients
> > >>> > >> > will
> > >>> > >> > > > fail
> > >>> > >> > > > > > if
> > >>> > >> > > > > > > they do not support zstandard.
> > >>> > >> > > > > > >
> > >>> > >> > > > > > > What do you think?
> > >>> > >> > > > > > >
> > >>> > >> > > > > > > Thanks,
> > >>> > >> > > > > > > Jason
> > >>> > >> > > > > > >
> > >>> > >> > > > > > > On Fri, Aug 17, 2018 at 8:08 AM, Dongjin Lee <
> > >>> > >> dongjin@apache.org
> > >>> > >> > >
> > >>> > >> > > > > wrote:
> > >>> > >> > > > > > >
> > >>> > >> > > > > > > > Thanks Jason, I reviewed the down-converting logic
> > >>> > following
> > >>> > >> > your
> > >>> > >> > > > > > > > explanation.[^1] You mean the following routines,
> > >>> right?
> > >>> > >> > > > > > > >
> > >>> > >> > > > > > > > -
> > >>> > >> > > > > > > >
> > https://github.com/apache/kafka/blob/trunk/core/src/
> > >>> > >> > > > > > > > main/scala/kafka/server/KafkaApis.scala#L534
> > >>> > >> > > > > > > > -
> > >>> > >> > > > > > > >
> > >>> https://github.com/apache/kafka/blob/trunk/clients/src/
> > >>> > >> > > > > > > > main/java/org/apache/kafka/common/record/
> > >>> > >> > > LazyDownConversionRecords.
> > >>> > >> > > > > > > > java#L165
> > >>> > >> > > > > > > > -
> > >>> > >> > > > > > > >
> > >>> https://github.com/apache/kafka/blob/trunk/clients/src/
> > >>> > >> > > > > > > >
> > >>> > >> main/java/org/apache/kafka/common/record/RecordsUtil.java#L40
> > >>> > >> > > > > > > >
> > >>> > >> > > > > > > > It seems like your stance is like following:
> > >>> > >> > > > > > > >
> > >>> > >> > > > > > > > 1. In principle, Kafka does not change the
> > compression
> > >>> > codec
> > >>> > >> > when
> > >>> > >> > > > > > > > down-converting, since it requires inspecting the
> > >>> fetched
> > >>> > >> data,
> > >>> > >> > > > which
> > >>> > >> > > > > > is
> > >>> > >> > > > > > > > expensive.
> > >>> > >> > > > > > > > 2. However, there are some cases the fetched data
> is
> > >>> > >> inspected
> > >>> > >> > > > > anyway.
> > >>> > >> > > > > > In
> > >>> > >> > > > > > > > this case, we can provide compression conversion
> > from
> > >>> > >> Zstandard
> > >>> > >> > > to
> > >>> > >> > > > > > > > classical ones[^2].
> > >>> > >> > > > > > > >
> > >>> > >> > > > > > > > And from what I understand, the cases where the
> > client
> > >>> > >> without
> > >>> > >> > > > > > ZStandard
> > >>> > >> > > > > > > > support receives ZStandard compressed records can
> be
> > >>> > >> organized
> > >>> > >> > > into
> > >>> > >> > > > > two
> > >>> > >> > > > > > > > cases:
> > >>> > >> > > > > > > >
> > >>> > >> > > > > > > > a. The 'compression.type' configuration of given
> > >>> topic is
> > >>> > >> > > > 'producer'
> > >>> > >> > > > > > and
> > >>> > >> > > > > > > > the producer compressed the records with
> ZStandard.
> > >>> (that
> > >>> > >> is,
> > >>> > >> > > using
> > >>> > >> > > > > > > > ZStandard implicitly.)
> > >>> > >> > > > > > > > b.  The 'compression.type' configuration of given
> > >>> topic is
> > >>> > >> > > 'zstd';
> > >>> > >> > > > > that
> > >>> > >> > > > > > > is,
> > >>> > >> > > > > > > > using ZStandard explicitly.
> > >>> > >> > > > > > > >
> > >>> > >> > > > > > > > As you stated, we don't have to handle the case b
> > >>> > specially.
> > >>> > >> > So,
> > >>> > >> > > It
> > >>> > >> > > > > > seems
> > >>> > >> > > > > > > > like we can narrow the focus of the problem by
> > joining
> > >>> > case
> > >>> > >> 1
> > >>> > >> > and
> > >>> > >> > > > > case
> > >>> > >> > > > > > b
> > >>> > >> > > > > > > > like the following:
> > >>> > >> > > > > > > >
> > >>> > >> > > > > > > > > Given the topic with 'producer' as its
> > >>> > 'compression.type'
> > >>> > >> > > > > > > configuration,
> > >>> > >> > > > > > > > ZStandard compressed records and old client
> without
> > >>> > >> ZStandard,
> > >>> > >> > is
> > >>> > >> > > > > there
> > >>> > >> > > > > > > any
> > >>> > >> > > > > > > > case we need to inspect the records and can change
> > the
> > >>> > >> > > compression
> > >>> > >> > > > > > type?
> > >>> > >> > > > > > > If
> > >>> > >> > > > > > > > so, can we provide compression type converting?
> > >>> > >> > > > > > > >
> > >>> > >> > > > > > > > Do I understand correctly?
> > >>> > >> > > > > > > >
> > >>> > >> > > > > > > > Best,
> > >>> > >> > > > > > > > Dongjin
> > >>> > >> > > > > > > >
> > >>> > >> > > > > > > > [^1]: I'm sorry, I found that I was a little bit
> > >>> > >> > misunderstanding
> > >>> > >> > > > how
> > >>> > >> > > > > > API
> > >>> > >> > > > > > > > version works, after reviewing the downconvert
> logic
> > >>> & the
> > >>> > >> > > protocol
> > >>> > >> > > > > > > > documentation <https://kafka.apache.org/protocol
> >.
> > >>> > >> > > > > > > > [^2]: None, Gzip, Snappy, Lz4.
> > >>> > >> > > > > > > >
> > >>> > >> > > > > > > > On Tue, Aug 14, 2018 at 2:16 AM Jason Gustafson <
> > >>> > >> > > > jason@confluent.io>
> > >>> > >> > > > > > > > wrote:
> > >>> > >> > > > > > > >
> > >>> > >> > > > > > > > > >
> > >>> > >> > > > > > > > > > But in my opinion, since the client will fail
> > >>> with the
> > >>> > >> API
> > >>> > >> > > > > version,
> > >>> > >> > > > > > > so
> > >>> > >> > > > > > > > we
> > >>> > >> > > > > > > > > > don't need to down-convert the messages
> anyway.
> > >>> Isn't
> > >>> > >> it?
> > >>> > >> > > So, I
> > >>> > >> > > > > > think
> > >>> > >> > > > > > > > we
> > >>> > >> > > > > > > > > > don't care about this case. (I'm sorry, I am
> not
> > >>> > >> familiar
> > >>> > >> > > with
> > >>> > >> > > > > > > > > down-convert
> > >>> > >> > > > > > > > > > logic.)
> > >>> > >> > > > > > > > >
> > >>> > >> > > > > > > > >
> > >>> > >> > > > > > > > > Currently the broker down-converts automatically
> > >>> when it
> > >>> > >> > > receives
> > >>> > >> > > > > an
> > >>> > >> > > > > > > old
> > >>> > >> > > > > > > > > version of the fetch request (a version which is
> > >>> known
> > >>> > to
> > >>> > >> > > predate
> > >>> > >> > > > > the
> > >>> > >> > > > > > > > > message format in use). Typically when
> > >>> down-converting
> > >>> > the
> > >>> > >> > > > message
> > >>> > >> > > > > > > > format,
> > >>> > >> > > > > > > > > we use the same compression type, but there is
> not
> > >>> much
> > >>> > >> point
> > >>> > >> > > in
> > >>> > >> > > > > > doing
> > >>> > >> > > > > > > so
> > >>> > >> > > > > > > > > when we know the client doesn't support it. So
> if
> > >>> > >> zstandard
> > >>> > >> > is
> > >>> > >> > > in
> > >>> > >> > > > > > use,
> > >>> > >> > > > > > > > and
> > >>> > >> > > > > > > > > we have to down-convert anyway, then we can
> choose
> > >>> to
> > >>> > use
> > >>> > >> a
> > >>> > >> > > > > different
> > >>> > >> > > > > > > > > compression type or no compression type.
> > >>> > >> > > > > > > > >
> > >>> > >> > > > > > > > > From my perspective, there is no significant
> > >>> downside to
> > >>> > >> > > bumping
> > >>> > >> > > > > the
> > >>> > >> > > > > > > > > protocol version and it has several potential
> > >>> benefits.
> > >>> > >> > Version
> > >>> > >> > > > > bumps
> > >>> > >> > > > > > > are
> > >>> > >> > > > > > > > > cheap. The main question mark in my mind is
> about
> > >>> > >> > > > down-conversion.
> > >>> > >> > > > > > > > Figuring
> > >>> > >> > > > > > > > > out whether down-conversion is needed is hard
> > >>> generally
> > >>> > >> > without
> > >>> > >> > > > > > > > inspecting
> > >>> > >> > > > > > > > > the fetched data, which is expensive. I think we
> > >>> agree
> > >>> > in
> > >>> > >> > > > principle
> > >>> > >> > > > > > > that
> > >>> > >> > > > > > > > we
> > >>> > >> > > > > > > > > do not want to have to pay this cost generally
> and
> > >>> > prefer
> > >>> > >> the
> > >>> > >> > > > > clients
> > >>> > >> > > > > > > to
> > >>> > >> > > > > > > > > fail when they see an unhandled compression
> type.
> > >>> The
> > >>> > >> point I
> > >>> > >> > > was
> > >>> > >> > > > > > > making
> > >>> > >> > > > > > > > is
> > >>> > >> > > > > > > > > that there are some cases where we are either
> > >>> inspecting
> > >>> > >> the
> > >>> > >> > > data
> > >>> > >> > > > > > > anyway
> > >>> > >> > > > > > > > > (because we have to down-convert the message
> > >>> format), or
> > >>> > >> we
> > >>> > >> > > have
> > >>> > >> > > > an
> > >>> > >> > > > > > > easy
> > >>> > >> > > > > > > > > way to tell whether zstandard is in use (the
> topic
> > >>> has
> > >>> > it
> > >>> > >> > > > > configured
> > >>> > >> > > > > > > > > explicitly). In the latter case, we don't have
> to
> > >>> handle
> > >>> > >> it
> > >>> > >> > > > > > specially.
> > >>> > >> > > > > > > > But
> > >>> > >> > > > > > > > > we do have to decide how we will handle
> > >>> down-conversion
> > >>> > to
> > >>> > >> > > older
> > >>> > >> > > > > > > formats.
> > >>> > >> > > > > > > > >
> > >>> > >> > > > > > > > > -Jason
> > >>> > >> > > > > > > > >
> > >>> > >> > > > > > > > > On Sun, Aug 12, 2018 at 5:15 PM, Dongjin Lee <
> > >>> > >> > > dongjin@apache.org
> > >>> > >> > > > >
> > >>> > >> > > > > > > wrote:
> > >>> > >> > > > > > > > >
> > >>> > >> > > > > > > > > > Colin and Jason,
> > >>> > >> > > > > > > > > >
> > >>> > >> > > > > > > > > > Thanks for your opinions. In summarizing, the
> > >>> Pros and
> > >>> > >> Cons
> > >>> > >> > > of
> > >>> > >> > > > > > > bumping
> > >>> > >> > > > > > > > > > fetch API version are:
> > >>> > >> > > > > > > > > >
> > >>> > >> > > > > > > > > > Cons:
> > >>> > >> > > > > > > > > >
> > >>> > >> > > > > > > > > > - The Broker can't know whether a given
> message
> > >>> batch
> > >>> > is
> > >>> > >> > > > > compressed
> > >>> > >> > > > > > > > with
> > >>> > >> > > > > > > > > > zstd or not.
> > >>> > >> > > > > > > > > > - Need some additional logic for the topic
> > >>> explicitly
> > >>> > >> > > > configured
> > >>> > >> > > > > to
> > >>> > >> > > > > > > use
> > >>> > >> > > > > > > > > > zstd.
> > >>> > >> > > > > > > > > >
> > >>> > >> > > > > > > > > > Pros:
> > >>> > >> > > > > > > > > >
> > >>> > >> > > > > > > > > > - The broker doesn't need to conduct expensive
> > >>> > >> > > down-conversion.
> > >>> > >> > > > > > > > > > - Can message the users to update their
> client.
> > >>> > >> > > > > > > > > >
> > >>> > >> > > > > > > > > > So, opinions for the backward-compatibility
> > >>> policy by
> > >>> > >> far:
> > >>> > >> > > > > > > > > >
> > >>> > >> > > > > > > > > > - A: bump the API version - +2 (Colin, Jason)
> > >>> > >> > > > > > > > > > - B: leave unchanged - +1 (Viktor)
> > >>> > >> > > > > > > > > >
> > >>> > >> > > > > > > > > > Here are my additional comments:
> > >>> > >> > > > > > > > > >
> > >>> > >> > > > > > > > > > @Colin
> > >>> > >> > > > > > > > > >
> > >>> > >> > > > > > > > > > I greatly appreciate your response. In the
> case
> > >>> of the
> > >>> > >> > > > dictionary
> > >>> > >> > > > > > > > > support,
> > >>> > >> > > > > > > > > > of course, this issue should be addressed
> later
> > >>> so we
> > >>> > >> don't
> > >>> > >> > > > need
> > >>> > >> > > > > it
> > >>> > >> > > > > > > in
> > >>> > >> > > > > > > > > the
> > >>> > >> > > > > > > > > > first version. You are right - it is not late
> to
> > >>> try
> > >>> > it
> > >>> > >> > after
> > >>> > >> > > > > some
> > >>> > >> > > > > > > > > > benchmarks. What I mean is, we should keep in
> > >>> mind on
> > >>> > >> that
> > >>> > >> > > > > > potential
> > >>> > >> > > > > > > > > > feature.
> > >>> > >> > > > > > > > > >
> > >>> > >> > > > > > > > > > @Jason
> > >>> > >> > > > > > > > > >
> > >>> > >> > > > > > > > > > You wrote,
> > >>> > >> > > > > > > > > >
> > >>> > >> > > > > > > > > > > Similarly, if we have to down-convert anyway
> > >>> because
> > >>> > >> the
> > >>> > >> > > > client
> > >>> > >> > > > > > > does
> > >>> > >> > > > > > > > > not
> > >>> > >> > > > > > > > > > understand the message format, then we could
> > also
> > >>> use
> > >>> > a
> > >>> > >> > > > different
> > >>> > >> > > > > > > > > > compression type.
> > >>> > >> > > > > > > > > >
> > >>> > >> > > > > > > > > > But in my opinion, since the client will fail
> > >>> with the
> > >>> > >> API
> > >>> > >> > > > > version,
> > >>> > >> > > > > > > so
> > >>> > >> > > > > > > > we
> > >>> > >> > > > > > > > > > don't need to down-convert the messages
> anyway.
> > >>> Isn't
> > >>> > >> it?
> > >>> > >> > > So, I
> > >>> > >> > > > > > think
> > >>> > >> > > > > > > > we
> > >>> > >> > > > > > > > > > don't care about this case. (I'm sorry, I am
> not
> > >>> > >> familiar
> > >>> > >> > > with
> > >>> > >> > > > > > > > > down-convert
> > >>> > >> > > > > > > > > > logic.)
> > >>> > >> > > > > > > > > >
> > >>> > >> > > > > > > > > > Please give more opinions. Thanks!
> > >>> > >> > > > > > > > > >
> > >>> > >> > > > > > > > > > - Dongjin
> > >>> > >> > > > > > > > > >
> > >>> > >> > > > > > > > > >
> > >>> > >> > > > > > > > > > On Wed, Aug 8, 2018 at 6:41 AM Jason
> Gustafson <
> > >>> > >> > > > > jason@confluent.io
> > >>> > >> > > > > > >
> > >>> > >> > > > > > > > > wrote:
> > >>> > >> > > > > > > > > >
> > >>> > >> > > > > > > > > > > Hey Colin,
> > >>> > >> > > > > > > > > > >
> > >>> > >> > > > > > > > > > > The problem for the fetch API is that the
> > broker
> > >>> > does
> > >>> > >> not
> > >>> > >> > > > > > generally
> > >>> > >> > > > > > > > > know
> > >>> > >> > > > > > > > > > if
> > >>> > >> > > > > > > > > > > a batch was compressed with zstd unless it
> > >>> parses
> > >>> > it.
> > >>> > >> I
> > >>> > >> > > think
> > >>> > >> > > > > the
> > >>> > >> > > > > > > > goal
> > >>> > >> > > > > > > > > > here
> > >>> > >> > > > > > > > > > > is to avoid the expensive down-conversion
> that
> > >>> is
> > >>> > >> needed
> > >>> > >> > to
> > >>> > >> > > > > > ensure
> > >>> > >> > > > > > > > > > > compatibility because it is only necessary
> if
> > >>> zstd
> > >>> > is
> > >>> > >> > > > actually
> > >>> > >> > > > > in
> > >>> > >> > > > > > > > use.
> > >>> > >> > > > > > > > > > But
> > >>> > >> > > > > > > > > > > as long as old clients can parse the message
> > >>> format,
> > >>> > >> they
> > >>> > >> > > > > should
> > >>> > >> > > > > > > get
> > >>> > >> > > > > > > > a
> > >>> > >> > > > > > > > > > > reasonable error if they see an unsupported
> > >>> > >> compression
> > >>> > >> > > type
> > >>> > >> > > > in
> > >>> > >> > > > > > the
> > >>> > >> > > > > > > > > > > attributes. Basically the onus is on users
> to
> > >>> ensure
> > >>> > >> that
> > >>> > >> > > > their
> > >>> > >> > > > > > > > > consumers
> > >>> > >> > > > > > > > > > > have been updated prior to using zstd. It
> > seems
> > >>> > like a
> > >>> > >> > > > > reasonable
> > >>> > >> > > > > > > > > > tradeoff
> > >>> > >> > > > > > > > > > > to me. There are a couple cases that might
> be
> > >>> worth
> > >>> > >> > > thinking
> > >>> > >> > > > > > > through:
> > >>> > >> > > > > > > > > > >
> > >>> > >> > > > > > > > > > > 1. If a topic is explicitly configured to
> use
> > >>> zstd,
> > >>> > >> then
> > >>> > >> > we
> > >>> > >> > > > > don't
> > >>> > >> > > > > > > > need
> > >>> > >> > > > > > > > > to
> > >>> > >> > > > > > > > > > > check the fetched data for the compression
> > type
> > >>> to
> > >>> > >> know
> > >>> > >> > if
> > >>> > >> > > we
> > >>> > >> > > > > > need
> > >>> > >> > > > > > > > > > > down-conversion. If we did bump the Fetch
> API
> > >>> > version,
> > >>> > >> > then
> > >>> > >> > > > we
> > >>> > >> > > > > > > could
> > >>> > >> > > > > > > > > > handle
> > >>> > >> > > > > > > > > > > this case by either down-converting using a
> > >>> > different
> > >>> > >> > > > > compression
> > >>> > >> > > > > > > > type
> > >>> > >> > > > > > > > > or
> > >>> > >> > > > > > > > > > > returning an error.
> > >>> > >> > > > > > > > > > > 2. Similarly, if we have to down-convert
> > anyway
> > >>> > >> because
> > >>> > >> > the
> > >>> > >> > > > > > client
> > >>> > >> > > > > > > > does
> > >>> > >> > > > > > > > > > not
> > >>> > >> > > > > > > > > > > understand the message format, then we could
> > >>> also
> > >>> > use
> > >>> > >> a
> > >>> > >> > > > > different
> > >>> > >> > > > > > > > > > > compression type.
> > >>> > >> > > > > > > > > > >
> > >>> > >> > > > > > > > > > > For the produce API, I think it's reasonable
> > to
> > >>> bump
> > >>> > >> the
> > >>> > >> > > api
> > >>> > >> > > > > > > version.
> > >>> > >> > > > > > > > > > This
> > >>> > >> > > > > > > > > > > can be used by clients to check whether a
> > broker
> > >>> > >> supports
> > >>> > >> > > > zstd.
> > >>> > >> > > > > > For
> > >>> > >> > > > > > > > > > > example, we might support a list of
> preferred
> > >>> > >> compression
> > >>> > >> > > > types
> > >>> > >> > > > > > in
> > >>> > >> > > > > > > > the
> > >>> > >> > > > > > > > > > > producer and we could use the broker to
> detect
> > >>> which
> > >>> > >> > > version
> > >>> > >> > > > to
> > >>> > >> > > > > > > use.
> > >>> > >> > > > > > > > > > >
> > >>> > >> > > > > > > > > > > -Jason
> > >>> > >> > > > > > > > > > >
> > >>> > >> > > > > > > > > > > On Tue, Aug 7, 2018 at 1:32 PM, Colin
> McCabe <
> > >>> > >> > > > > cmccabe@apache.org
> > >>> > >> > > > > > >
> > >>> > >> > > > > > > > > wrote:
> > >>> > >> > > > > > > > > > >
> > >>> > >> > > > > > > > > > > > Thanks for bumping this, Dongjin.  ZStd
> is a
> > >>> good
> > >>> > >> > > > compression
> > >>> > >> > > > > > > codec
> > >>> > >> > > > > > > > > > and I
> > >>> > >> > > > > > > > > > > > hope we can get this support in soon!
> > >>> > >> > > > > > > > > > > >
> > >>> > >> > > > > > > > > > > > I would say we can just bump the API
> version
> > >>> to
> > >>> > >> > indicate
> > >>> > >> > > > that
> > >>> > >> > > > > > > ZStd
> > >>> > >> > > > > > > > > > > support
> > >>> > >> > > > > > > > > > > > is expected in new clients.  We probably
> > need
> > >>> some
> > >>> > >> way
> > >>> > >> > of
> > >>> > >> > > > > > > > indicating
> > >>> > >> > > > > > > > > to
> > >>> > >> > > > > > > > > > > the
> > >>> > >> > > > > > > > > > > > older clients that they can't consume the
> > >>> > >> partitions,
> > >>> > >> > as
> > >>> > >> > > > > well.
> > >>> > >> > > > > > > > > Perhaps
> > >>> > >> > > > > > > > > > > we
> > >>> > >> > > > > > > > > > > > can use the UNSUPPORTED_FOR_MESSAGE_FORMAT
> > >>> error?
> > >>> > >> > > > > > > > > > > >
> > >>> > >> > > > > > > > > > > > The license thing seems straightforward --
> > >>> it's
> > >>> > >> just a
> > >>> > >> > > > matter
> > >>> > >> > > > > > of
> > >>> > >> > > > > > > > > adding
> > >>> > >> > > > > > > > > > > > the text to the right files as per ASF
> > >>> guidelines.
> > >>> > >> > > > > > > > > > > >
> > >>> > >> > > > > > > > > > > > With regard to the dictionary support, do
> we
> > >>> > really
> > >>> > >> > need
> > >>> > >> > > > that
> > >>> > >> > > > > > in
> > >>> > >> > > > > > > > the
> > >>> > >> > > > > > > > > > > first
> > >>> > >> > > > > > > > > > > > version?  Hopefully message batches are
> big
> > >>> enough
> > >>> > >> that
> > >>> > >> > > > this
> > >>> > >> > > > > > > isn't
> > >>> > >> > > > > > > > > > > needed.
> > >>> > >> > > > > > > > > > > > Some benchmarks might help here.
> > >>> > >> > > > > > > > > > > >
> > >>> > >> > > > > > > > > > > > best,
> > >>> > >> > > > > > > > > > > > Colin
> > >>> > >> > > > > > > > > > > >
> > >>> > >> > > > > > > > > > > >
> > >>> > >> > > > > > > > > > > > On Tue, Aug 7, 2018, at 08:02, Dongjin Lee
> > >>> wrote:
> > >>> > >> > > > > > > > > > > > > As Kafka 2.0.0 was released, let's
> reboot
> > >>> this
> > >>> > >> issue,
> > >>> > >> > > > > KIP-110
> > >>> > >> > > > > > > > > > > > > <https://cwiki.apache.org/
> > >>> > >> > > confluence/display/KAFKA/KIP-
> > >>> > >> > > > > > > > > > > > 110%3A+Add+Codec+for+
> ZStandard+Compression>
> > >>> > >> > > > > > > > > > > > > .
> > >>> > >> > > > > > > > > > > > >
> > >>> > >> > > > > > > > > > > > > For newcomers, Here is some summary of
> the
> > >>> > >> history:
> > >>> > >> > > > KIP-110
> > >>> > >> > > > > > was
> > >>> > >> > > > > > > > > > > > originally
> > >>> > >> > > > > > > > > > > > > worked for the issue KAFKA-4514 but, it
> > >>> lacked
> > >>> > >> > > benchmark
> > >>> > >> > > > > > > results
> > >>> > >> > > > > > > > to
> > >>> > >> > > > > > > > > > get
> > >>> > >> > > > > > > > > > > > the
> > >>> > >> > > > > > > > > > > > > agreement of the community. Later, Ivan
> > >>> Babrou
> > >>> > and
> > >>> > >> > some
> > >>> > >> > > > > other
> > >>> > >> > > > > > > > users
> > >>> > >> > > > > > > > > > who
> > >>> > >> > > > > > > > > > > > > adopted the patch provided their
> excellent
> > >>> > >> > performance
> > >>> > >> > > > > report
> > >>> > >> > > > > > > > which
> > >>> > >> > > > > > > > > > is
> > >>> > >> > > > > > > > > > > > now
> > >>> > >> > > > > > > > > > > > > included in the KIP, but it postponed
> > again
> > >>> > >> because
> > >>> > >> > of
> > >>> > >> > > > the
> > >>> > >> > > > > > > > > community
> > >>> > >> > > > > > > > > > > was
> > >>> > >> > > > > > > > > > > > > busy for 2.0.0 release. It is why I now
> > >>> reboot
> > >>> > >> this
> > >>> > >> > > > issue.
> > >>> > >> > > > > > > > > > > > >
> > >>> > >> > > > > > > > > > > > > The following is the current status of
> the
> > >>> > >> feature:
> > >>> > >> > You
> > >>> > >> > > > can
> > >>> > >> > > > > > > check
> > >>> > >> > > > > > > > > the
> > >>> > >> > > > > > > > > > > > > current draft implementation here
> > >>> > >> > > > > > > > > > > > > <
> > https://github.com/apache/kafka/pull/2267>.
> > >>> It
> > >>> > >> is
> > >>> > >> > > based
> > >>> > >> > > > > on
> > >>> > >> > > > > > > zstd
> > >>> > >> > > > > > > > > > 1.3.5
> > >>> > >> > > > > > > > > > > > and
> > >>> > >> > > > > > > > > > > > > periodically rebased onto the latest
> > >>> trunk[^1].
> > >>> > >> > > > > > > > > > > > >
> > >>> > >> > > > > > > > > > > > > The issues that should be addressed is
> > like
> > >>> > >> > following:
> > >>> > >> > > > > > > > > > > > >
> > >>> > >> > > > > > > > > > > > > *1. Backward Compatibility*
> > >>> > >> > > > > > > > > > > > >
> > >>> > >> > > > > > > > > > > > > To support old consumers, we need to
> take
> > a
> > >>> > >> strategy
> > >>> > >> > to
> > >>> > >> > > > > > handle
> > >>> > >> > > > > > > > the
> > >>> > >> > > > > > > > > > old
> > >>> > >> > > > > > > > > > > > > consumers. Current candidates are:
> > >>> > >> > > > > > > > > > > > >
> > >>> > >> > > > > > > > > > > > > - Bump API version
> > >>> > >> > > > > > > > > > > > > - Leave unchanged: let the old clients
> > fail.
> > >>> > >> > > > > > > > > > > > > - Improve the error messages:
> > >>> > >> > > > > > > > > > > > >
> > >>> > >> > > > > > > > > > > > > *2. Dictionary Support*
> > >>> > >> > > > > > > > > > > > >
> > >>> > >> > > > > > > > > > > > > To support zstd's dictionary feature in
> > the
> > >>> > future
> > >>> > >> > (if
> > >>> > >> > > > > > needed),
> > >>> > >> > > > > > > > we
> > >>> > >> > > > > > > > > > need
> > >>> > >> > > > > > > > > > > > to
> > >>> > >> > > > > > > > > > > > > sketch how it should be and leave some
> > room
> > >>> for
> > >>> > >> it.
> > >>> > >> > As
> > >>> > >> > > of
> > >>> > >> > > > > > now,
> > >>> > >> > > > > > > > > there
> > >>> > >> > > > > > > > > > > has
> > >>> > >> > > > > > > > > > > > > been no discussion on this topic yet.
> > >>> > >> > > > > > > > > > > > >
> > >>> > >> > > > > > > > > > > > > *3. License*
> > >>> > >> > > > > > > > > > > > >
> > >>> > >> > > > > > > > > > > > > To use this feature, we need to add
> > license
> > >>> of
> > >>> > >> zstd
> > >>> > >> > and
> > >>> > >> > > > > > > zstd-jni
> > >>> > >> > > > > > > > to
> > >>> > >> > > > > > > > > > the
> > >>> > >> > > > > > > > > > > > > project. (Thanks to Viktor Somogyi for
> > >>> raising
> > >>> > >> this
> > >>> > >> > > > issue!)
> > >>> > >> > > > > > It
> > >>> > >> > > > > > > > > seems
> > >>> > >> > > > > > > > > > > like
> > >>> > >> > > > > > > > > > > > > what Apache Spark did would be a good
> > >>> example
> > >>> > but
> > >>> > >> > there
> > >>> > >> > > > has
> > >>> > >> > > > > > > been
> > >>> > >> > > > > > > > no
> > >>> > >> > > > > > > > > > > > > discussion yet.
> > >>> > >> > > > > > > > > > > > >
> > >>> > >> > > > > > > > > > > > > You can find the details of the above
> > >>> issues in
> > >>> > >> the
> > >>> > >> > KIP
> > >>> > >> > > > > > > document.
> > >>> > >> > > > > > > > > > > Please
> > >>> > >> > > > > > > > > > > > > have a look when you are free, and give
> me
> > >>> > >> feedback.
> > >>> > >> > > All
> > >>> > >> > > > > > kinds
> > >>> > >> > > > > > > of
> > >>> > >> > > > > > > > > > > > > participating are welcome.
> > >>> > >> > > > > > > > > > > > >
> > >>> > >> > > > > > > > > > > > > Best,
> > >>> > >> > > > > > > > > > > > > Dongjin
> > >>> > >> > > > > > > > > > > > >
> > >>> > >> > > > > > > > > > > > > [^1]: At the time of writing, commit
> > >>> 6b4fb8152.
> > >>> > >> > > > > > > > > > > > >
> > >>> > >> > > > > > > > > > > > > On Sat, Jul 14, 2018 at 10:45 PM Dongjin
> > >>> Lee <
> > >>> > >> > > > > > > dongjin@apache.org
> > >>> > >> > > > > > > > >
> > >>> > >> > > > > > > > > > > wrote:
> > >>> > >> > > > > > > > > > > > >
> > >>> > >> > > > > > > > > > > > > > Sorry for the late reply.
> > >>> > >> > > > > > > > > > > > > >
> > >>> > >> > > > > > > > > > > > > > In short, I could not submit the
> updated
> > >>> KIP
> > >>> > by
> > >>> > >> the
> > >>> > >> > > > > feature
> > >>> > >> > > > > > > > > freeze
> > >>> > >> > > > > > > > > > > > > > deadline of 2.0.0. For this reason, it
> > >>> will
> > >>> > not
> > >>> > >> be
> > >>> > >> > > > > included
> > >>> > >> > > > > > > in
> > >>> > >> > > > > > > > > the
> > >>> > >> > > > > > > > > > > > 2.0.0
> > >>> > >> > > > > > > > > > > > > > release and all discussion for this
> > issue
> > >>> were
> > >>> > >> > > > postponed
> > >>> > >> > > > > > > after
> > >>> > >> > > > > > > > > the
> > >>> > >> > > > > > > > > > > > release
> > >>> > >> > > > > > > > > > > > > > of 2.0.0.
> > >>> > >> > > > > > > > > > > > > >
> > >>> > >> > > > > > > > > > > > > > I have been updating the PR following
> > >>> recent
> > >>> > >> > updates.
> > >>> > >> > > > > Just
> > >>> > >> > > > > > > > now, I
> > >>> > >> > > > > > > > > > > > rebased
> > >>> > >> > > > > > > > > > > > > > it against the latest trunk and
> updated
> > >>> the
> > >>> > zstd
> > >>> > >> > > > version
> > >>> > >> > > > > > into
> > >>> > >> > > > > > > > > > 1.3.5.
> > >>> > >> > > > > > > > > > > > If you
> > >>> > >> > > > > > > > > > > > > > need some request, don't hesitate to
> > >>> notify
> > >>> > me.
> > >>> > >> > (But
> > >>> > >> > > > not
> > >>> > >> > > > > > this
> > >>> > >> > > > > > > > > > thread
> > >>> > >> > > > > > > > > > > -
> > >>> > >> > > > > > > > > > > > just
> > >>> > >> > > > > > > > > > > > > > send me the message directly.)
> > >>> > >> > > > > > > > > > > > > >
> > >>> > >> > > > > > > > > > > > > > Best,
> > >>> > >> > > > > > > > > > > > > > Dongjin
> > >>> > >> > > > > > > > > > > > > >
> > >>> > >> > > > > > > > > > > > > > On Tue, Jul 10, 2018 at 11:57 PM Bobby
> > >>> Evans <
> > >>> > >> > > > > > > bobby@apache.org
> > >>> > >> > > > > > > > >
> > >>> > >> > > > > > > > > > > wrote:
> > >>> > >> > > > > > > > > > > > > >
> > >>> > >> > > > > > > > > > > > > >> I there any update on this.  The
> > >>> performance
> > >>> > >> > > > > improvements
> > >>> > >> > > > > > > are
> > >>> > >> > > > > > > > > > quite
> > >>> > >> > > > > > > > > > > > > >> impressive and I really would like to
> > >>> stop
> > >>> > >> forking
> > >>> > >> > > > kafka
> > >>> > >> > > > > > > just
> > >>> > >> > > > > > > > to
> > >>> > >> > > > > > > > > > get
> > >>> > >> > > > > > > > > > > > this
> > >>> > >> > > > > > > > > > > > > >> in.
> > >>> > >> > > > > > > > > > > > > >>
> > >>> > >> > > > > > > > > > > > > >> Thanks,
> > >>> > >> > > > > > > > > > > > > >>
> > >>> > >> > > > > > > > > > > > > >> Bobby
> > >>> > >> > > > > > > > > > > > > >>
> > >>> > >> > > > > > > > > > > > > >> On Wed, Jun 13, 2018 at 8:56 PM
> Dongjin
> > >>> Lee <
> > >>> > >> > > > > > > > dongjin@apache.org
> > >>> > >> > > > > > > > > >
> > >>> > >> > > > > > > > > > > > wrote:
> > >>> > >> > > > > > > > > > > > > >>
> > >>> > >> > > > > > > > > > > > > >> > Ismael,
> > >>> > >> > > > > > > > > > > > > >> >
> > >>> > >> > > > > > > > > > > > > >> > Oh, I forgot all of you are on
> > working
> > >>> > frenzy
> > >>> > >> > for
> > >>> > >> > > > 2.0!
> > >>> > >> > > > > > No
> > >>> > >> > > > > > > > > > problem,
> > >>> > >> > > > > > > > > > > > take
> > >>> > >> > > > > > > > > > > > > >> > your time. I am also working at
> > another
> > >>> > issue
> > >>> > >> > now.
> > >>> > >> > > > > Thank
> > >>> > >> > > > > > > you
> > >>> > >> > > > > > > > > for
> > >>> > >> > > > > > > > > > > > > >> letting me
> > >>> > >> > > > > > > > > > > > > >> > know.
> > >>> > >> > > > > > > > > > > > > >> >
> > >>> > >> > > > > > > > > > > > > >> > Best,
> > >>> > >> > > > > > > > > > > > > >> > Dongjin
> > >>> > >> > > > > > > > > > > > > >> >
> > >>> > >> > > > > > > > > > > > > >> > On Wed, Jun 13, 2018, 11:44 PM
> Ismael
> > >>> Juma
> > >>> > <
> > >>> > >> > > > > > > > ismaelj@gmail.com
> > >>> > >> > > > > > > > > >
> > >>> > >> > > > > > > > > > > > wrote:
> > >>> > >> > > > > > > > > > > > > >> >
> > >>> > >> > > > > > > > > > > > > >> > > Sorry for the delay Dongjin.
> > >>> Everyone is
> > >>> > >> busy
> > >>> > >> > > > > > finalising
> > >>> > >> > > > > > > > > > 2.0.0.
> > >>> > >> > > > > > > > > > > > This
> > >>> > >> > > > > > > > > > > > > >> KIP
> > >>> > >> > > > > > > > > > > > > >> > > seems like a great candidate for
> > >>> 2.1.0
> > >>> > and
> > >>> > >> > > > hopefully
> > >>> > >> > > > > > > there
> > >>> > >> > > > > > > > > > will
> > >>> > >> > > > > > > > > > > be
> > >>> > >> > > > > > > > > > > > > >> more
> > >>> > >> > > > > > > > > > > > > >> > of
> > >>> > >> > > > > > > > > > > > > >> > > a discussion next week. :)
> > >>> > >> > > > > > > > > > > > > >> > >
> > >>> > >> > > > > > > > > > > > > >> > > Ismael
> > >>> > >> > > > > > > > > > > > > >> > >
> > >>> > >> > > > > > > > > > > > > >> > > On Wed, 13 Jun 2018, 05:17
> Dongjin
> > >>> Lee, <
> > >>> > >> > > > > > > > dongjin@apache.org
> > >>> > >> > > > > > > > > >
> > >>> > >> > > > > > > > > > > > wrote:
> > >>> > >> > > > > > > > > > > > > >> > >
> > >>> > >> > > > > > > > > > > > > >> > > > Hello. I just updated my draft
> > >>> > >> > implementation:
> > >>> > >> > > > > > > > > > > > > >> > > >
> > >>> > >> > > > > > > > > > > > > >> > > > 1. Rebased to latest trunk
> > (commit
> > >>> > >> 5145d6b)
> > >>> > >> > > > > > > > > > > > > >> > > > 2. Apply ZStd 1.3.4
> > >>> > >> > > > > > > > > > > > > >> > > >
> > >>> > >> > > > > > > > > > > > > >> > > > You can check out the
> > >>> implementation
> > >>> > from
> > >>> > >> > here
> > >>> > >> > > > > > > > > > > > > >> > > > <
> > >>> > >> https://github.com/apache/kafka/pull/2267
> > >>> > >> > >.
> > >>> > >> > > If
> > >>> > >> > > > > you
> > >>> > >> > > > > > > > > > > experience
> > >>> > >> > > > > > > > > > > > any
> > >>> > >> > > > > > > > > > > > > >> > > problem
> > >>> > >> > > > > > > > > > > > > >> > > > running it, don't hesitate to
> > give
> > >>> me a
> > >>> > >> > > mention.
> > >>> > >> > > > > > > > > > > > > >> > > >
> > >>> > >> > > > > > > > > > > > > >> > > > Best,
> > >>> > >> > > > > > > > > > > > > >> > > > Dongjin
> > >>> > >> > > > > > > > > > > > > >> > > >
> > >>> > >> > > > > > > > > > > > > >> > > > On Tue, Jun 12, 2018 at 6:50 PM
> > >>> Dongjin
> > >>> > >> Lee
> > >>> > >> > <
> > >>> > >> > > > > > > > > > > dongjin@apache.org
> > >>> > >> > > > > > > > > > > > >
> > >>> > >> > > > > > > > > > > > > >> > wrote:
> > >>> > >> > > > > > > > > > > > > >> > > >
> > >>> > >> > > > > > > > > > > > > >> > > > > Here is the short conclusion
> > >>> about
> > >>> > the
> > >>> > >> > > license
> > >>> > >> > > > > > > > problem:
> > >>> > >> > > > > > > > > > *We
> > >>> > >> > > > > > > > > > > > can
> > >>> > >> > > > > > > > > > > > > >> use
> > >>> > >> > > > > > > > > > > > > >> > > zstd
> > >>> > >> > > > > > > > > > > > > >> > > > > and zstd-jni without any
> > >>> problem, but
> > >>> > >> we
> > >>> > >> > > need
> > >>> > >> > > > to
> > >>> > >> > > > > > > > include
> > >>> > >> > > > > > > > > > > their
> > >>> > >> > > > > > > > > > > > > >> > license,
> > >>> > >> > > > > > > > > > > > > >> > > > > e.g., BSD license.*
> > >>> > >> > > > > > > > > > > > > >> > > > >
> > >>> > >> > > > > > > > > > > > > >> > > > > Both of BSD 2 Clause License
> &
> > 3
> > >>> > Clause
> > >>> > >> > > > License
> > >>> > >> > > > > > > > requires
> > >>> > >> > > > > > > > > > to
> > >>> > >> > > > > > > > > > > > > >> include
> > >>> > >> > > > > > > > > > > > > >> > the
> > >>> > >> > > > > > > > > > > > > >> > > > > license used, and BSD 3
> Clause
> > >>> > License
> > >>> > >> > > > requires
> > >>> > >> > > > > > that
> > >>> > >> > > > > > > > the
> > >>> > >> > > > > > > > > > > name
> > >>> > >> > > > > > > > > > > > of
> > >>> > >> > > > > > > > > > > > > >> the
> > >>> > >> > > > > > > > > > > > > >> > > > > contributor can't be used to
> > >>> endorse
> > >>> > or
> > >>> > >> > > > promote
> > >>> > >> > > > > > the
> > >>> > >> > > > > > > > > > product.
> > >>> > >> > > > > > > > > > > > > >> That's
> > >>> > >> > > > > > > > > > > > > >> > it
> > >>> > >> > > > > > > > > > > > > >> > > > > <
> > >>> > >> > > > > > > > > > > > > >> > > >
> > >>> > >> > > > > > > > > > > > > >> > >
> > >>> > >> > > > > > > > > > > > > >> >
> > >>> > >> > > > > > > > > > > > > >>
> > >>> > >> > http://www.mikestratton.net/2011/12/is-bsd-license-
> > >>> > >> > > > > > > > > > > > compatible-with-apache-2-0-license/
> > >>> > >> > > > > > > > > > > > > >> > > > >
> > >>> > >> > > > > > > > > > > > > >> > > > > - They are not listed in the
> > >>> list of
> > >>> > >> > > > prohibited
> > >>> > >> > > > > > > > licenses
> > >>> > >> > > > > > > > > > > > > >> > > > > <
> https://www.apache.org/legal/
> > >>> > >> > > > > > > > resolved.html#category-x>
> > >>> > >> > > > > > > > > > > also.
> > >>> > >> > > > > > > > > > > > > >> > > > >
> > >>> > >> > > > > > > > > > > > > >> > > > > Here is how Spark did for it
> > >>> > >> > > > > > > > > > > > > >> > > > > <https://issues.apache.org/
> > >>> > >> > > > > > jira/browse/SPARK-19112
> > >>> > >> > > > > > > >:
> > >>> > >> > > > > > > > > > > > > >> > > > >
> > >>> > >> > > > > > > > > > > > > >> > > > > - They made a directory
> > >>> dedicated to
> > >>> > >> the
> > >>> > >> > > > > > dependency
> > >>> > >> > > > > > > > > > license
> > >>> > >> > > > > > > > > > > > files
> > >>> > >> > > > > > > > > > > > > >> > > > > <
> > >>> > >> > > > > > > https://github.com/apache/
> spark/tree/master/licenses
> > >>> > >> > > > > > > > >
> > >>> > >> > > > > > > > > > and
> > >>> > >> > > > > > > > > > > > added
> > >>> > >> > > > > > > > > > > > > >> > > > licenses
> > >>> > >> > > > > > > > > > > > > >> > > > > for Zstd
> > >>> > >> > > > > > > > > > > > > >> > > > > <
> > >>> > >> > > > > > > > > > > > > >> >
> > >>> > >> > > > https://github.com/apache/spark/blob/master/licenses/
> > >>> > >> > > > > > > > > > > > LICENSE-zstd.txt
> > >>> > >> > > > > > > > > > > > > >> > > >
> > >>> > >> > > > > > > > > > > > > >> > > > &
> > >>> > >> > > > > > > > > > > > > >> > > > > Zstd-jni
> > >>> > >> > > > > > > > > > > > > >> > > > > <
> > >>> > >> > > > > > > > > > > > > >> > > >
> > >>> > >> > > > > > > > > > > > > >> > >
> > >>> > >> > > > > > > > > > > > > >> >
> > >>> > >> > > > > > > > > > > > > >> https://github.com/apache/
> > >>> > >> > > spark/blob/master/licenses/
> > >>> > >> > > > > > > > > > > > LICENSE-zstd-jni.txt
> > >>> > >> > > > > > > > > > > > > >> >
> > >>> > >> > > > > > > > > > > > > >> > > > > .
> > >>> > >> > > > > > > > > > > > > >> > > > > - Added a link to the
> original
> > >>> > license
> > >>> > >> > files
> > >>> > >> > > > in
> > >>> > >> > > > > > > > LICENSE.
> > >>> > >> > > > > > > > > > > > > >> > > > > <
> > >>> > >> > > > > https://github.com/apache/spark/pull/18805/files
> > >>> > >> > > > > > >
> > >>> > >> > > > > > > > > > > > > >> > > > >
> > >>> > >> > > > > > > > > > > > > >> > > > > If needed, I can make a
> similar
> > >>> > update.
> > >>> > >> > > > > > > > > > > > > >> > > > >
> > >>> > >> > > > > > > > > > > > > >> > > > > Thanks for pointing out this
> > >>> problem,
> > >>> > >> > > Viktor!
> > >>> > >> > > > > Nice
> > >>> > >> > > > > > > > > catch!
> > >>> > >> > > > > > > > > > > > > >> > > > >
> > >>> > >> > > > > > > > > > > > > >> > > > > Best,
> > >>> > >> > > > > > > > > > > > > >> > > > > Dongjin
> > >>> > >> > > > > > > > > > > > > >> > > > >
> > >>> > >> > > > > > > > > > > > > >> > > > >
> > >>> > >> > > > > > > > > > > > > >> > > > >
> > >>> > >> > > > > > > > > > > > > >> > > > > On Mon, Jun 11, 2018 at 11:50
> > PM
> > >>> > >> Dongjin
> > >>> > >> > > Lee <
> > >>> > >> > > > > > > > > > > > dongjin@apache.org>
> > >>> > >> > > > > > > > > > > > > >> > > wrote:
> > >>> > >> > > > > > > > > > > > > >> > > > >
> > >>> > >> > > > > > > > > > > > > >> > > > >> I greatly appreciate your
> > >>> > >> comprehensive
> > >>> > >> > > > > > reasoning.
> > >>> > >> > > > > > > > so:
> > >>> > >> > > > > > > > > +1
> > >>> > >> > > > > > > > > > > > for b
> > >>> > >> > > > > > > > > > > > > >> > until
> > >>> > >> > > > > > > > > > > > > >> > > > now.
> > >>> >
>

Re: [DISCUSS] KIP-110: Add Codec for ZStandard Compression (Updated)

[Magnus Edenhill <ma...@edenhill.se>]

> Ismael wrote:
> Jason, that's an interesting point regarding the Java client. Do we know
> what clients in other languages do in these cases?

librdkafka (and its bindings) passes unknown/future errors through to the
application, the error code remains intact while
the error string will be set to something like "Err-123?", which isn't very
helpful to the user but it at least
preserves the original error code for further troubleshooting.
For the producer any unknown error returned in the ProduceResponse will be
considered a permanent delivery failure (no retries),
and for the consumer any unknown FetchResponse errors will propagate
directly to the application, trigger a fetch backoff, and then
continue fetching past that offset.

So, from the client's perspective it is not really a problem if new error
codes are added to older API versions.

/Magnus


Den tors 6 sep. 2018 kl 09:45 skrev Dongjin Lee <do...@apache.org>:

> I updated the KIP page
> <
> https://cwiki.apache.org/confluence/display/KAFKA/KIP-110%3A+Add+Codec+for+ZStandard+Compression
> >
> following the discussion here. Please take a look when you are free.
> If you have any opinion, don't hesitate to give me a message.
>
> Best,
> Dongjin
>
> On Fri, Aug 31, 2018 at 11:35 PM Dongjin Lee <do...@apache.org> wrote:
>
> > I just updated the draft implementation[^1], rebasing against the latest
> > trunk and implementing error routine (i.e., Error code 74 for
> > UnsupportedCompressionTypeException.) Since we decided to disallow all
> > fetch request below version 2.1.0 for the topics specifying ZStandard, I
> > added an error logic only.
> >
> > Please have a look when you are free.
> >
> > Thanks,
> > Dongjin
> >
> > [^1]: Please check the last commit here:
> > https://github.com/apache/kafka/pull/2267
> >
> > On Thu, Aug 23, 2018, 8:55 AM Dongjin Lee <do...@apache.org> wrote:
> >
> >> Jason,
> >>
> >> Great. +1 for UNSUPPORTED_COMPRESSION_TYPE.
> >>
> >> Best,
> >> Dongjin
> >>
> >> On Thu, Aug 23, 2018 at 8:19 AM Jason Gustafson <ja...@confluent.io>
> >> wrote:
> >>
> >>> Hey Dongjin,
> >>>
> >>> Yeah that's right. For what it's worth, librdkafka also appears to
> handle
> >>> unexpected error codes. I expect that most client implementations would
> >>> either pass through the raw type or convert to an enum using something
> >>> like
> >>> what the java client does. Since we're expecting the client to fail
> >>> anyway,
> >>> I'm probably in favor of using the UNSUPPORTED_COMPRESSION_TYPE error
> >>> code.
> >>>
> >>> -Jason
> >>>
> >>> On Wed, Aug 22, 2018 at 1:46 AM, Dongjin Lee <do...@apache.org>
> wrote:
> >>>
> >>> > Jason and Ismael,
> >>> >
> >>> > It seems like the only thing we need to regard if we define a new
> error
> >>> > code (i.e., UNSUPPORTED_COMPRESSION_TYPE) would be the implementation
> >>> of
> >>> > the other language clients, right? At least, this strategy causes any
> >>> > problem for Java client. Do I understand correctly?
> >>> >
> >>> > Thanks,
> >>> > Dongjin
> >>> >
> >>> > On Wed, Aug 22, 2018 at 5:43 PM Dongjin Lee <do...@apache.org>
> >>> wrote:
> >>> >
> >>> > > Jason,
> >>> > >
> >>> > > > I think we would only use this error code when we /know/ that
> zstd
> >>> was
> >>> > > in use and the client doesn't support it? This is true if either 1)
> >>> the
> >>> > > message needs down-conversion and we encounter a zstd compressed
> >>> message,
> >>> > > or 2) if the topic is explicitly configured to use zstd.
> >>> > >
> >>> > > Yes, it is right. And you know, the case 1 includes 1.a) old
> clients'
> >>> > > request v0, v1 records or 1.b) implicit zstd, the compression type
> of
> >>> > > "producer" with Zstd compressed data.
> >>> > >
> >>> > > > However, if the compression type is set to "producer," then the
> >>> fetched
> >>> > > data may or may not be compressed with zstd. In this case, we
> return
> >>> the
> >>> > > data to the client and expect it to fail parsing. Is that correct?
> >>> > >
> >>> > > Exactly.
> >>> > >
> >>> > > Following your message, I reviewed the implementation of
> >>> > > `KafkaApis#handleFetchRequest,` which handles the fetch request.
> And
> >>> > found
> >>> > > that the information we can use is like the following:
> >>> > >
> >>> > > 1. Client's fetch request version. (`versionId` variable)
> >>> > > 2. Log's compression type. (`logConfig` variable)
> >>> > >
> >>> > > We can't detect the actual compression type of the data, unless we
> >>> > inspect
> >>> > > the `RecordBatch` included in the `Records` instance (i.e.,
> >>> > > `unconvertedRecords` variable.) Since it requires some performance
> >>> issue,
> >>> > > it is not our option - in short, we can't be sure if given chunks
> of
> >>> data
> >>> > > are compressed with zstd or not.
> >>> > >
> >>> > > So, conclusion: we can return an error in the case of 1.a and 2
> >>> easily,
> >>> > > with the information above. In the case 1.b (implicit zstd), we can
> >>> just
> >>> > > return the data by do nothing special and expect it to fail
> parsing.
> >>> > >
> >>> > > Thanks,
> >>> > > Dongjin
> >>> > >
> >>> > > On Wed, Aug 22, 2018 at 12:02 PM Ismael Juma <is...@gmail.com>
> >>> wrote:
> >>> > >
> >>> > >> Jason, that's an interesting point regarding the Java client. Do
> we
> >>> know
> >>> > >> what clients in other languages do in these cases?
> >>> > >>
> >>> > >> Ismael
> >>> > >>
> >>> > >> On Tue, 21 Aug 2018, 17:30 Jason Gustafson, <ja...@confluent.io>
> >>> wrote:
> >>> > >>
> >>> > >> > Hi Dongjin,
> >>> > >> >
> >>> > >> > One of the complications is that old versions of the API will
> not
> >>> > >> expect a
> >>> > >> > new error code. However, since we expect this to be a fatal
> error
> >>> > anyway
> >>> > >> > for old clients, it may still be more useful to return the
> correct
> >>> > error
> >>> > >> > code. For example, the Kafka clients use the following code to
> >>> convert
> >>> > >> the
> >>> > >> > error code:
> >>> > >> >
> >>> > >> >     public static Errors forCode(short code) {
> >>> > >> >         Errors error = codeToError.get(code);
> >>> > >> >         if (error != null) {
> >>> > >> >             return error;
> >>> > >> >         } else {
> >>> > >> >             log.warn("Unexpected error code: {}.", code);
> >>> > >> >             return UNKNOWN_SERVER_ERROR;
> >>> > >> >         }
> >>> > >> >     }
> >>> > >> >
> >>> > >> > If we return an unsupported error code, it will be converted to
> an
> >>> > >> UNKNOWN
> >>> > >> > error, but at least we will get the message in the log with the
> >>> > correct
> >>> > >> > code. That seems preferable to returning a misleading error
> code.
> >>> So I
> >>> > >> > wonder if we can use the new UNSUPPORTED_COMPRESSION_TYPE error
> >>> even
> >>> > for
> >>> > >> > older versions.
> >>> > >> >
> >>> > >> > Also, one question just to check my understanding. I think we
> >>> would
> >>> > only
> >>> > >> > use this error code when we /know/ that zstd was in use and the
> >>> client
> >>> > >> > doesn't support it? This is true if either 1) the message needs
> >>> > >> > down-conversion and we encounter a zstd compressed message, or
> 2)
> >>> if
> >>> > the
> >>> > >> > topic is explicitly configured to use zstd. However, if the
> >>> > compression
> >>> > >> > type is set to "producer," then the fetched data may or may not
> be
> >>> > >> > compressed with zstd. In this case, we return the data to the
> >>> client
> >>> > and
> >>> > >> > expect it to fail parsing. Is that correct?
> >>> > >> >
> >>> > >> > Thanks,
> >>> > >> > Jason
> >>> > >> >
> >>> > >> >
> >>> > >> >
> >>> > >> > On Tue, Aug 21, 2018 at 9:08 AM, Dongjin Lee <
> dongjin@apache.org>
> >>> > >> wrote:
> >>> > >> >
> >>> > >> > > Ismael, Jason and all,
> >>> > >> > >
> >>> > >> > > I rewrote the backward compatibility strategy & its
> alternatives
> >>> > like
> >>> > >> > > following, based on Ismael & Jason's comments. Since it is not
> >>> > >> updated to
> >>> > >> > > the wiki yet, don't hesitate to give me a message if you have
> >>> any
> >>> > >> opinion
> >>> > >> > > on it.
> >>> > >> > >
> >>> > >> > > ```
> >>> > >> > > *Backward Compatibility*
> >>> > >> > >
> >>> > >> > > We need to establish some backward-compatibility strategy for
> >>> the
> >>> > >> case an
> >>> > >> > > old client subscribes a topic using ZStandard implicitly
> (i.e.,
> >>> > >> > > 'compression.type' configuration of given topic is 'producer'
> >>> and
> >>> > the
> >>> > >> > > producer compressed the records with ZStandard). We have the
> >>> > following
> >>> > >> > > options for this situation:
> >>> > >> > >
> >>> > >> > > *A. Support ZStandard to the old clients which can understand
> >>> v0, v1
> >>> > >> > > messages only.*
> >>> > >> > >
> >>> > >> > > This strategy necessarily requires the down-conversion of v2
> >>> message
> >>> > >> > > compressed with Zstandard into v0 or v1 messages, which means
> a
> >>> > >> > > considerable performance degradation. So we rejected this
> >>> strategy.
> >>> > >> > >
> >>> > >> > > *B. Bump the API version and support only v2-available
> clients*
> >>> > >> > >
> >>> > >> > > With this approach, we can message the old clients that they
> >>> are old
> >>> > >> and
> >>> > >> > > should be upgraded. However, there are still several options
> >>> for the
> >>> > >> > Error
> >>> > >> > > code.
> >>> > >> > >
> >>> > >> > > *B.1. INVALID_REQUEST (42)*
> >>> > >> > >
> >>> > >> > > This option gives the client so little information; the user
> >>> can be
> >>> > >> > > confused about why the client worked correctly in the past
> >>> suddenly
> >>> > >> > > encounters a problem. So we rejected this strategy.
> >>> > >> > >
> >>> > >> > > *B.2. CORRUPT_MESSAGE (2)*
> >>> > >> > >
> >>> > >> > > This option gives inaccurate information; the user can be
> >>> surprised
> >>> > >> and
> >>> > >> > > misunderstand that the log files are broken in some way. So we
> >>> > >> rejected
> >>> > >> > > this strategy.
> >>> > >> > >
> >>> > >> > > *B.3 UNSUPPORTED_FOR_MESSAGE_FORMAT (43)*
> >>> > >> > >
> >>> > >> > > The advantage of this approach is we don't need to define a
> new
> >>> > error
> >>> > >> > code;
> >>> > >> > > we can reuse it and that's all.
> >>> > >> > >
> >>> > >> > > The disadvantage of this approach is that it is also a little
> >>> bit
> >>> > >> vague;
> >>> > >> > > This error code is defined as a work for KIP-98[^1] and now
> >>> returned
> >>> > >> in
> >>> > >> > the
> >>> > >> > > transaction error.
> >>> > >> > >
> >>> > >> > > *B.4. UNSUPPORTED_COMPRESSION_TYPE (new)*
> >>> > >> > >
> >>> > >> > > The advantage of this approach is that it is clear and
> provides
> >>> an
> >>> > >> exact
> >>> > >> > > description. The disadvantage is we need to add a new error
> >>> code.
> >>> > >> > > ```
> >>> > >> > >
> >>> > >> > > *It seems like what we need to choose is now so clear:
> >>> > >> > > UNSUPPORTED_FOR_MESSAGE_FORMAT (B.3) or
> >>> UNSUPPORTED_COMPRESSION_TYPE
> >>> > >> > > (B.4).*
> >>> > >> > > The first one doesn't need a new error message but the latter
> is
> >>> > more
> >>> > >> > > explicit. Which one do you prefer? Since all of you have much
> >>> more
> >>> > >> > > experience and knowledge than me, I will follow your decision.
> >>> The
> >>> > >> wiki
> >>> > >> > > page will be updated following the decision also.
> >>> > >> > >
> >>> > >> > > Best,
> >>> > >> > > Dongjin
> >>> > >> > >
> >>> > >> > > [^1]: https://issues.apache.org/jira/browse/KAFKA-4990
> >>> > >> > >
> >>> > >> > > On Sun, Aug 19, 2018 at 4:58 AM Ismael Juma <
> ismaelj@gmail.com>
> >>> > >> wrote:
> >>> > >> > >
> >>> > >> > > > Sounds reasonable to me.
> >>> > >> > > >
> >>> > >> > > > Ismael
> >>> > >> > > >
> >>> > >> > > > On Sat, 18 Aug 2018, 12:20 Jason Gustafson, <
> >>> jason@confluent.io>
> >>> > >> > wrote:
> >>> > >> > > >
> >>> > >> > > > > Hey Ismael,
> >>> > >> > > > >
> >>> > >> > > > > Your summary looks good to me. I think it might also be a
> >>> good
> >>> > >> idea
> >>> > >> > to
> >>> > >> > > > add
> >>> > >> > > > > a new UNSUPPORTED_COMPRESSION_TYPE error code to go along
> >>> with
> >>> > the
> >>> > >> > > > version
> >>> > >> > > > > bumps. We won't be able to use it for old api versions
> >>> since the
> >>> > >> > > clients
> >>> > >> > > > > will not understand it, but we can use it going forward so
> >>> that
> >>> > >> we're
> >>> > >> > > not
> >>> > >> > > > > stuck in a similar situation with a new message format and
> >>> a new
> >>> > >> > codec
> >>> > >> > > to
> >>> > >> > > > > support. Another option is to use UNSUPPORTED_FOR_MESSAGE_
> >>> > FORMAT,
> >>> > >> but
> >>> > >> > > it
> >>> > >> > > > is
> >>> > >> > > > > not as explicit.
> >>> > >> > > > >
> >>> > >> > > > > -Jason
> >>> > >> > > > >
> >>> > >> > > > > On Fri, Aug 17, 2018 at 5:19 PM, Ismael Juma <
> >>> ismael@juma.me.uk
> >>> > >
> >>> > >> > > wrote:
> >>> > >> > > > >
> >>> > >> > > > > > Hi Dongjin and Jason,
> >>> > >> > > > > >
> >>> > >> > > > > > I would agree. My summary:
> >>> > >> > > > > >
> >>> > >> > > > > > 1. Support zstd with message format 2 only.
> >>> > >> > > > > > 2. Bump produce and fetch request versions.
> >>> > >> > > > > > 3. Provide broker errors whenever possible based on the
> >>> > request
> >>> > >> > > version
> >>> > >> > > > > and
> >>> > >> > > > > > rely on clients for the cases where the broker can't
> >>> validate
> >>> > >> > > > efficiently
> >>> > >> > > > > > (example message format 2 consumer that supports the
> >>> latest
> >>> > >> fetch
> >>> > >> > > > version
> >>> > >> > > > > > but doesn't support zstd).
> >>> > >> > > > > >
> >>> > >> > > > > > If there's general agreement on this, I suggest we
> update
> >>> the
> >>> > >> KIP
> >>> > >> > to
> >>> > >> > > > > state
> >>> > >> > > > > > the proposal and to move the rejected options to its own
> >>> > >> section.
> >>> > >> > And
> >>> > >> > > > > then
> >>> > >> > > > > > start a vote!
> >>> > >> > > > > >
> >>> > >> > > > > > Ismael
> >>> > >> > > > > >
> >>> > >> > > > > > On Fri, Aug 17, 2018 at 4:00 PM Jason Gustafson <
> >>> > >> > jason@confluent.io>
> >>> > >> > > > > > wrote:
> >>> > >> > > > > >
> >>> > >> > > > > > > Hi Dongjin,
> >>> > >> > > > > > >
> >>> > >> > > > > > > Yes, that's a good summary. For clients which support
> >>> v2,
> >>> > the
> >>> > >> > > client
> >>> > >> > > > > can
> >>> > >> > > > > > > parse the message format and hopefully raise a useful
> >>> error
> >>> > >> > message
> >>> > >> > > > > > > indicating the unsupported compression type. For older
> >>> > >> clients,
> >>> > >> > our
> >>> > >> > > > > > options
> >>> > >> > > > > > > are probably (1) to down-convert to the old format
> >>> using no
> >>> > >> > > > compression
> >>> > >> > > > > > > type, or (2) to return an error code. I'm leaning
> >>> toward the
> >>> > >> > latter
> >>> > >> > > > as
> >>> > >> > > > > > the
> >>> > >> > > > > > > simpler solution, but the challenge is finding a good
> >>> error
> >>> > >> code.
> >>> > >> > > Two
> >>> > >> > > > > > > possibilities might be INVALID_REQUEST or
> >>> CORRUPT_MESSAGE.
> >>> > The
> >>> > >> > > > downside
> >>> > >> > > > > > is
> >>> > >> > > > > > > that old clients probably won't get a helpful message.
> >>> > >> However,
> >>> > >> > at
> >>> > >> > > > > least
> >>> > >> > > > > > > the behavior will be consistent in the sense that all
> >>> > clients
> >>> > >> > will
> >>> > >> > > > fail
> >>> > >> > > > > > if
> >>> > >> > > > > > > they do not support zstandard.
> >>> > >> > > > > > >
> >>> > >> > > > > > > What do you think?
> >>> > >> > > > > > >
> >>> > >> > > > > > > Thanks,
> >>> > >> > > > > > > Jason
> >>> > >> > > > > > >
> >>> > >> > > > > > > On Fri, Aug 17, 2018 at 8:08 AM, Dongjin Lee <
> >>> > >> dongjin@apache.org
> >>> > >> > >
> >>> > >> > > > > wrote:
> >>> > >> > > > > > >
> >>> > >> > > > > > > > Thanks Jason, I reviewed the down-converting logic
> >>> > following
> >>> > >> > your
> >>> > >> > > > > > > > explanation.[^1] You mean the following routines,
> >>> right?
> >>> > >> > > > > > > >
> >>> > >> > > > > > > > -
> >>> > >> > > > > > > >
> https://github.com/apache/kafka/blob/trunk/core/src/
> >>> > >> > > > > > > > main/scala/kafka/server/KafkaApis.scala#L534
> >>> > >> > > > > > > > -
> >>> > >> > > > > > > >
> >>> https://github.com/apache/kafka/blob/trunk/clients/src/
> >>> > >> > > > > > > > main/java/org/apache/kafka/common/record/
> >>> > >> > > LazyDownConversionRecords.
> >>> > >> > > > > > > > java#L165
> >>> > >> > > > > > > > -
> >>> > >> > > > > > > >
> >>> https://github.com/apache/kafka/blob/trunk/clients/src/
> >>> > >> > > > > > > >
> >>> > >> main/java/org/apache/kafka/common/record/RecordsUtil.java#L40
> >>> > >> > > > > > > >
> >>> > >> > > > > > > > It seems like your stance is like following:
> >>> > >> > > > > > > >
> >>> > >> > > > > > > > 1. In principle, Kafka does not change the
> compression
> >>> > codec
> >>> > >> > when
> >>> > >> > > > > > > > down-converting, since it requires inspecting the
> >>> fetched
> >>> > >> data,
> >>> > >> > > > which
> >>> > >> > > > > > is
> >>> > >> > > > > > > > expensive.
> >>> > >> > > > > > > > 2. However, there are some cases the fetched data is
> >>> > >> inspected
> >>> > >> > > > > anyway.
> >>> > >> > > > > > In
> >>> > >> > > > > > > > this case, we can provide compression conversion
> from
> >>> > >> Zstandard
> >>> > >> > > to
> >>> > >> > > > > > > > classical ones[^2].
> >>> > >> > > > > > > >
> >>> > >> > > > > > > > And from what I understand, the cases where the
> client
> >>> > >> without
> >>> > >> > > > > > ZStandard
> >>> > >> > > > > > > > support receives ZStandard compressed records can be
> >>> > >> organized
> >>> > >> > > into
> >>> > >> > > > > two
> >>> > >> > > > > > > > cases:
> >>> > >> > > > > > > >
> >>> > >> > > > > > > > a. The 'compression.type' configuration of given
> >>> topic is
> >>> > >> > > > 'producer'
> >>> > >> > > > > > and
> >>> > >> > > > > > > > the producer compressed the records with ZStandard.
> >>> (that
> >>> > >> is,
> >>> > >> > > using
> >>> > >> > > > > > > > ZStandard implicitly.)
> >>> > >> > > > > > > > b.  The 'compression.type' configuration of given
> >>> topic is
> >>> > >> > > 'zstd';
> >>> > >> > > > > that
> >>> > >> > > > > > > is,
> >>> > >> > > > > > > > using ZStandard explicitly.
> >>> > >> > > > > > > >
> >>> > >> > > > > > > > As you stated, we don't have to handle the case b
> >>> > specially.
> >>> > >> > So,
> >>> > >> > > It
> >>> > >> > > > > > seems
> >>> > >> > > > > > > > like we can narrow the focus of the problem by
> joining
> >>> > case
> >>> > >> 1
> >>> > >> > and
> >>> > >> > > > > case
> >>> > >> > > > > > b
> >>> > >> > > > > > > > like the following:
> >>> > >> > > > > > > >
> >>> > >> > > > > > > > > Given the topic with 'producer' as its
> >>> > 'compression.type'
> >>> > >> > > > > > > configuration,
> >>> > >> > > > > > > > ZStandard compressed records and old client without
> >>> > >> ZStandard,
> >>> > >> > is
> >>> > >> > > > > there
> >>> > >> > > > > > > any
> >>> > >> > > > > > > > case we need to inspect the records and can change
> the
> >>> > >> > > compression
> >>> > >> > > > > > type?
> >>> > >> > > > > > > If
> >>> > >> > > > > > > > so, can we provide compression type converting?
> >>> > >> > > > > > > >
> >>> > >> > > > > > > > Do I understand correctly?
> >>> > >> > > > > > > >
> >>> > >> > > > > > > > Best,
> >>> > >> > > > > > > > Dongjin
> >>> > >> > > > > > > >
> >>> > >> > > > > > > > [^1]: I'm sorry, I found that I was a little bit
> >>> > >> > misunderstanding
> >>> > >> > > > how
> >>> > >> > > > > > API
> >>> > >> > > > > > > > version works, after reviewing the downconvert logic
> >>> & the
> >>> > >> > > protocol
> >>> > >> > > > > > > > documentation <https://kafka.apache.org/protocol>.
> >>> > >> > > > > > > > [^2]: None, Gzip, Snappy, Lz4.
> >>> > >> > > > > > > >
> >>> > >> > > > > > > > On Tue, Aug 14, 2018 at 2:16 AM Jason Gustafson <
> >>> > >> > > > jason@confluent.io>
> >>> > >> > > > > > > > wrote:
> >>> > >> > > > > > > >
> >>> > >> > > > > > > > > >
> >>> > >> > > > > > > > > > But in my opinion, since the client will fail
> >>> with the
> >>> > >> API
> >>> > >> > > > > version,
> >>> > >> > > > > > > so
> >>> > >> > > > > > > > we
> >>> > >> > > > > > > > > > don't need to down-convert the messages anyway.
> >>> Isn't
> >>> > >> it?
> >>> > >> > > So, I
> >>> > >> > > > > > think
> >>> > >> > > > > > > > we
> >>> > >> > > > > > > > > > don't care about this case. (I'm sorry, I am not
> >>> > >> familiar
> >>> > >> > > with
> >>> > >> > > > > > > > > down-convert
> >>> > >> > > > > > > > > > logic.)
> >>> > >> > > > > > > > >
> >>> > >> > > > > > > > >
> >>> > >> > > > > > > > > Currently the broker down-converts automatically
> >>> when it
> >>> > >> > > receives
> >>> > >> > > > > an
> >>> > >> > > > > > > old
> >>> > >> > > > > > > > > version of the fetch request (a version which is
> >>> known
> >>> > to
> >>> > >> > > predate
> >>> > >> > > > > the
> >>> > >> > > > > > > > > message format in use). Typically when
> >>> down-converting
> >>> > the
> >>> > >> > > > message
> >>> > >> > > > > > > > format,
> >>> > >> > > > > > > > > we use the same compression type, but there is not
> >>> much
> >>> > >> point
> >>> > >> > > in
> >>> > >> > > > > > doing
> >>> > >> > > > > > > so
> >>> > >> > > > > > > > > when we know the client doesn't support it. So if
> >>> > >> zstandard
> >>> > >> > is
> >>> > >> > > in
> >>> > >> > > > > > use,
> >>> > >> > > > > > > > and
> >>> > >> > > > > > > > > we have to down-convert anyway, then we can choose
> >>> to
> >>> > use
> >>> > >> a
> >>> > >> > > > > different
> >>> > >> > > > > > > > > compression type or no compression type.
> >>> > >> > > > > > > > >
> >>> > >> > > > > > > > > From my perspective, there is no significant
> >>> downside to
> >>> > >> > > bumping
> >>> > >> > > > > the
> >>> > >> > > > > > > > > protocol version and it has several potential
> >>> benefits.
> >>> > >> > Version
> >>> > >> > > > > bumps
> >>> > >> > > > > > > are
> >>> > >> > > > > > > > > cheap. The main question mark in my mind is about
> >>> > >> > > > down-conversion.
> >>> > >> > > > > > > > Figuring
> >>> > >> > > > > > > > > out whether down-conversion is needed is hard
> >>> generally
> >>> > >> > without
> >>> > >> > > > > > > > inspecting
> >>> > >> > > > > > > > > the fetched data, which is expensive. I think we
> >>> agree
> >>> > in
> >>> > >> > > > principle
> >>> > >> > > > > > > that
> >>> > >> > > > > > > > we
> >>> > >> > > > > > > > > do not want to have to pay this cost generally and
> >>> > prefer
> >>> > >> the
> >>> > >> > > > > clients
> >>> > >> > > > > > > to
> >>> > >> > > > > > > > > fail when they see an unhandled compression type.
> >>> The
> >>> > >> point I
> >>> > >> > > was
> >>> > >> > > > > > > making
> >>> > >> > > > > > > > is
> >>> > >> > > > > > > > > that there are some cases where we are either
> >>> inspecting
> >>> > >> the
> >>> > >> > > data
> >>> > >> > > > > > > anyway
> >>> > >> > > > > > > > > (because we have to down-convert the message
> >>> format), or
> >>> > >> we
> >>> > >> > > have
> >>> > >> > > > an
> >>> > >> > > > > > > easy
> >>> > >> > > > > > > > > way to tell whether zstandard is in use (the topic
> >>> has
> >>> > it
> >>> > >> > > > > configured
> >>> > >> > > > > > > > > explicitly). In the latter case, we don't have to
> >>> handle
> >>> > >> it
> >>> > >> > > > > > specially.
> >>> > >> > > > > > > > But
> >>> > >> > > > > > > > > we do have to decide how we will handle
> >>> down-conversion
> >>> > to
> >>> > >> > > older
> >>> > >> > > > > > > formats.
> >>> > >> > > > > > > > >
> >>> > >> > > > > > > > > -Jason
> >>> > >> > > > > > > > >
> >>> > >> > > > > > > > > On Sun, Aug 12, 2018 at 5:15 PM, Dongjin Lee <
> >>> > >> > > dongjin@apache.org
> >>> > >> > > > >
> >>> > >> > > > > > > wrote:
> >>> > >> > > > > > > > >
> >>> > >> > > > > > > > > > Colin and Jason,
> >>> > >> > > > > > > > > >
> >>> > >> > > > > > > > > > Thanks for your opinions. In summarizing, the
> >>> Pros and
> >>> > >> Cons
> >>> > >> > > of
> >>> > >> > > > > > > bumping
> >>> > >> > > > > > > > > > fetch API version are:
> >>> > >> > > > > > > > > >
> >>> > >> > > > > > > > > > Cons:
> >>> > >> > > > > > > > > >
> >>> > >> > > > > > > > > > - The Broker can't know whether a given message
> >>> batch
> >>> > is
> >>> > >> > > > > compressed
> >>> > >> > > > > > > > with
> >>> > >> > > > > > > > > > zstd or not.
> >>> > >> > > > > > > > > > - Need some additional logic for the topic
> >>> explicitly
> >>> > >> > > > configured
> >>> > >> > > > > to
> >>> > >> > > > > > > use
> >>> > >> > > > > > > > > > zstd.
> >>> > >> > > > > > > > > >
> >>> > >> > > > > > > > > > Pros:
> >>> > >> > > > > > > > > >
> >>> > >> > > > > > > > > > - The broker doesn't need to conduct expensive
> >>> > >> > > down-conversion.
> >>> > >> > > > > > > > > > - Can message the users to update their client.
> >>> > >> > > > > > > > > >
> >>> > >> > > > > > > > > > So, opinions for the backward-compatibility
> >>> policy by
> >>> > >> far:
> >>> > >> > > > > > > > > >
> >>> > >> > > > > > > > > > - A: bump the API version - +2 (Colin, Jason)
> >>> > >> > > > > > > > > > - B: leave unchanged - +1 (Viktor)
> >>> > >> > > > > > > > > >
> >>> > >> > > > > > > > > > Here are my additional comments:
> >>> > >> > > > > > > > > >
> >>> > >> > > > > > > > > > @Colin
> >>> > >> > > > > > > > > >
> >>> > >> > > > > > > > > > I greatly appreciate your response. In the case
> >>> of the
> >>> > >> > > > dictionary
> >>> > >> > > > > > > > > support,
> >>> > >> > > > > > > > > > of course, this issue should be addressed later
> >>> so we
> >>> > >> don't
> >>> > >> > > > need
> >>> > >> > > > > it
> >>> > >> > > > > > > in
> >>> > >> > > > > > > > > the
> >>> > >> > > > > > > > > > first version. You are right - it is not late to
> >>> try
> >>> > it
> >>> > >> > after
> >>> > >> > > > > some
> >>> > >> > > > > > > > > > benchmarks. What I mean is, we should keep in
> >>> mind on
> >>> > >> that
> >>> > >> > > > > > potential
> >>> > >> > > > > > > > > > feature.
> >>> > >> > > > > > > > > >
> >>> > >> > > > > > > > > > @Jason
> >>> > >> > > > > > > > > >
> >>> > >> > > > > > > > > > You wrote,
> >>> > >> > > > > > > > > >
> >>> > >> > > > > > > > > > > Similarly, if we have to down-convert anyway
> >>> because
> >>> > >> the
> >>> > >> > > > client
> >>> > >> > > > > > > does
> >>> > >> > > > > > > > > not
> >>> > >> > > > > > > > > > understand the message format, then we could
> also
> >>> use
> >>> > a
> >>> > >> > > > different
> >>> > >> > > > > > > > > > compression type.
> >>> > >> > > > > > > > > >
> >>> > >> > > > > > > > > > But in my opinion, since the client will fail
> >>> with the
> >>> > >> API
> >>> > >> > > > > version,
> >>> > >> > > > > > > so
> >>> > >> > > > > > > > we
> >>> > >> > > > > > > > > > don't need to down-convert the messages anyway.
> >>> Isn't
> >>> > >> it?
> >>> > >> > > So, I
> >>> > >> > > > > > think
> >>> > >> > > > > > > > we
> >>> > >> > > > > > > > > > don't care about this case. (I'm sorry, I am not
> >>> > >> familiar
> >>> > >> > > with
> >>> > >> > > > > > > > > down-convert
> >>> > >> > > > > > > > > > logic.)
> >>> > >> > > > > > > > > >
> >>> > >> > > > > > > > > > Please give more opinions. Thanks!
> >>> > >> > > > > > > > > >
> >>> > >> > > > > > > > > > - Dongjin
> >>> > >> > > > > > > > > >
> >>> > >> > > > > > > > > >
> >>> > >> > > > > > > > > > On Wed, Aug 8, 2018 at 6:41 AM Jason Gustafson <
> >>> > >> > > > > jason@confluent.io
> >>> > >> > > > > > >
> >>> > >> > > > > > > > > wrote:
> >>> > >> > > > > > > > > >
> >>> > >> > > > > > > > > > > Hey Colin,
> >>> > >> > > > > > > > > > >
> >>> > >> > > > > > > > > > > The problem for the fetch API is that the
> broker
> >>> > does
> >>> > >> not
> >>> > >> > > > > > generally
> >>> > >> > > > > > > > > know
> >>> > >> > > > > > > > > > if
> >>> > >> > > > > > > > > > > a batch was compressed with zstd unless it
> >>> parses
> >>> > it.
> >>> > >> I
> >>> > >> > > think
> >>> > >> > > > > the
> >>> > >> > > > > > > > goal
> >>> > >> > > > > > > > > > here
> >>> > >> > > > > > > > > > > is to avoid the expensive down-conversion that
> >>> is
> >>> > >> needed
> >>> > >> > to
> >>> > >> > > > > > ensure
> >>> > >> > > > > > > > > > > compatibility because it is only necessary if
> >>> zstd
> >>> > is
> >>> > >> > > > actually
> >>> > >> > > > > in
> >>> > >> > > > > > > > use.
> >>> > >> > > > > > > > > > But
> >>> > >> > > > > > > > > > > as long as old clients can parse the message
> >>> format,
> >>> > >> they
> >>> > >> > > > > should
> >>> > >> > > > > > > get
> >>> > >> > > > > > > > a
> >>> > >> > > > > > > > > > > reasonable error if they see an unsupported
> >>> > >> compression
> >>> > >> > > type
> >>> > >> > > > in
> >>> > >> > > > > > the
> >>> > >> > > > > > > > > > > attributes. Basically the onus is on users to
> >>> ensure
> >>> > >> that
> >>> > >> > > > their
> >>> > >> > > > > > > > > consumers
> >>> > >> > > > > > > > > > > have been updated prior to using zstd. It
> seems
> >>> > like a
> >>> > >> > > > > reasonable
> >>> > >> > > > > > > > > > tradeoff
> >>> > >> > > > > > > > > > > to me. There are a couple cases that might be
> >>> worth
> >>> > >> > > thinking
> >>> > >> > > > > > > through:
> >>> > >> > > > > > > > > > >
> >>> > >> > > > > > > > > > > 1. If a topic is explicitly configured to use
> >>> zstd,
> >>> > >> then
> >>> > >> > we
> >>> > >> > > > > don't
> >>> > >> > > > > > > > need
> >>> > >> > > > > > > > > to
> >>> > >> > > > > > > > > > > check the fetched data for the compression
> type
> >>> to
> >>> > >> know
> >>> > >> > if
> >>> > >> > > we
> >>> > >> > > > > > need
> >>> > >> > > > > > > > > > > down-conversion. If we did bump the Fetch API
> >>> > version,
> >>> > >> > then
> >>> > >> > > > we
> >>> > >> > > > > > > could
> >>> > >> > > > > > > > > > handle
> >>> > >> > > > > > > > > > > this case by either down-converting using a
> >>> > different
> >>> > >> > > > > compression
> >>> > >> > > > > > > > type
> >>> > >> > > > > > > > > or
> >>> > >> > > > > > > > > > > returning an error.
> >>> > >> > > > > > > > > > > 2. Similarly, if we have to down-convert
> anyway
> >>> > >> because
> >>> > >> > the
> >>> > >> > > > > > client
> >>> > >> > > > > > > > does
> >>> > >> > > > > > > > > > not
> >>> > >> > > > > > > > > > > understand the message format, then we could
> >>> also
> >>> > use
> >>> > >> a
> >>> > >> > > > > different
> >>> > >> > > > > > > > > > > compression type.
> >>> > >> > > > > > > > > > >
> >>> > >> > > > > > > > > > > For the produce API, I think it's reasonable
> to
> >>> bump
> >>> > >> the
> >>> > >> > > api
> >>> > >> > > > > > > version.
> >>> > >> > > > > > > > > > This
> >>> > >> > > > > > > > > > > can be used by clients to check whether a
> broker
> >>> > >> supports
> >>> > >> > > > zstd.
> >>> > >> > > > > > For
> >>> > >> > > > > > > > > > > example, we might support a list of preferred
> >>> > >> compression
> >>> > >> > > > types
> >>> > >> > > > > > in
> >>> > >> > > > > > > > the
> >>> > >> > > > > > > > > > > producer and we could use the broker to detect
> >>> which
> >>> > >> > > version
> >>> > >> > > > to
> >>> > >> > > > > > > use.
> >>> > >> > > > > > > > > > >
> >>> > >> > > > > > > > > > > -Jason
> >>> > >> > > > > > > > > > >
> >>> > >> > > > > > > > > > > On Tue, Aug 7, 2018 at 1:32 PM, Colin McCabe <
> >>> > >> > > > > cmccabe@apache.org
> >>> > >> > > > > > >
> >>> > >> > > > > > > > > wrote:
> >>> > >> > > > > > > > > > >
> >>> > >> > > > > > > > > > > > Thanks for bumping this, Dongjin.  ZStd is a
> >>> good
> >>> > >> > > > compression
> >>> > >> > > > > > > codec
> >>> > >> > > > > > > > > > and I
> >>> > >> > > > > > > > > > > > hope we can get this support in soon!
> >>> > >> > > > > > > > > > > >
> >>> > >> > > > > > > > > > > > I would say we can just bump the API version
> >>> to
> >>> > >> > indicate
> >>> > >> > > > that
> >>> > >> > > > > > > ZStd
> >>> > >> > > > > > > > > > > support
> >>> > >> > > > > > > > > > > > is expected in new clients.  We probably
> need
> >>> some
> >>> > >> way
> >>> > >> > of
> >>> > >> > > > > > > > indicating
> >>> > >> > > > > > > > > to
> >>> > >> > > > > > > > > > > the
> >>> > >> > > > > > > > > > > > older clients that they can't consume the
> >>> > >> partitions,
> >>> > >> > as
> >>> > >> > > > > well.
> >>> > >> > > > > > > > > Perhaps
> >>> > >> > > > > > > > > > > we
> >>> > >> > > > > > > > > > > > can use the UNSUPPORTED_FOR_MESSAGE_FORMAT
> >>> error?
> >>> > >> > > > > > > > > > > >
> >>> > >> > > > > > > > > > > > The license thing seems straightforward --
> >>> it's
> >>> > >> just a
> >>> > >> > > > matter
> >>> > >> > > > > > of
> >>> > >> > > > > > > > > adding
> >>> > >> > > > > > > > > > > > the text to the right files as per ASF
> >>> guidelines.
> >>> > >> > > > > > > > > > > >
> >>> > >> > > > > > > > > > > > With regard to the dictionary support, do we
> >>> > really
> >>> > >> > need
> >>> > >> > > > that
> >>> > >> > > > > > in
> >>> > >> > > > > > > > the
> >>> > >> > > > > > > > > > > first
> >>> > >> > > > > > > > > > > > version?  Hopefully message batches are big
> >>> enough
> >>> > >> that
> >>> > >> > > > this
> >>> > >> > > > > > > isn't
> >>> > >> > > > > > > > > > > needed.
> >>> > >> > > > > > > > > > > > Some benchmarks might help here.
> >>> > >> > > > > > > > > > > >
> >>> > >> > > > > > > > > > > > best,
> >>> > >> > > > > > > > > > > > Colin
> >>> > >> > > > > > > > > > > >
> >>> > >> > > > > > > > > > > >
> >>> > >> > > > > > > > > > > > On Tue, Aug 7, 2018, at 08:02, Dongjin Lee
> >>> wrote:
> >>> > >> > > > > > > > > > > > > As Kafka 2.0.0 was released, let's reboot
> >>> this
> >>> > >> issue,
> >>> > >> > > > > KIP-110
> >>> > >> > > > > > > > > > > > > <https://cwiki.apache.org/
> >>> > >> > > confluence/display/KAFKA/KIP-
> >>> > >> > > > > > > > > > > > 110%3A+Add+Codec+for+ZStandard+Compression>
> >>> > >> > > > > > > > > > > > > .
> >>> > >> > > > > > > > > > > > >
> >>> > >> > > > > > > > > > > > > For newcomers, Here is some summary of the
> >>> > >> history:
> >>> > >> > > > KIP-110
> >>> > >> > > > > > was
> >>> > >> > > > > > > > > > > > originally
> >>> > >> > > > > > > > > > > > > worked for the issue KAFKA-4514 but, it
> >>> lacked
> >>> > >> > > benchmark
> >>> > >> > > > > > > results
> >>> > >> > > > > > > > to
> >>> > >> > > > > > > > > > get
> >>> > >> > > > > > > > > > > > the
> >>> > >> > > > > > > > > > > > > agreement of the community. Later, Ivan
> >>> Babrou
> >>> > and
> >>> > >> > some
> >>> > >> > > > > other
> >>> > >> > > > > > > > users
> >>> > >> > > > > > > > > > who
> >>> > >> > > > > > > > > > > > > adopted the patch provided their excellent
> >>> > >> > performance
> >>> > >> > > > > report
> >>> > >> > > > > > > > which
> >>> > >> > > > > > > > > > is
> >>> > >> > > > > > > > > > > > now
> >>> > >> > > > > > > > > > > > > included in the KIP, but it postponed
> again
> >>> > >> because
> >>> > >> > of
> >>> > >> > > > the
> >>> > >> > > > > > > > > community
> >>> > >> > > > > > > > > > > was
> >>> > >> > > > > > > > > > > > > busy for 2.0.0 release. It is why I now
> >>> reboot
> >>> > >> this
> >>> > >> > > > issue.
> >>> > >> > > > > > > > > > > > >
> >>> > >> > > > > > > > > > > > > The following is the current status of the
> >>> > >> feature:
> >>> > >> > You
> >>> > >> > > > can
> >>> > >> > > > > > > check
> >>> > >> > > > > > > > > the
> >>> > >> > > > > > > > > > > > > current draft implementation here
> >>> > >> > > > > > > > > > > > > <
> https://github.com/apache/kafka/pull/2267>.
> >>> It
> >>> > >> is
> >>> > >> > > based
> >>> > >> > > > > on
> >>> > >> > > > > > > zstd
> >>> > >> > > > > > > > > > 1.3.5
> >>> > >> > > > > > > > > > > > and
> >>> > >> > > > > > > > > > > > > periodically rebased onto the latest
> >>> trunk[^1].
> >>> > >> > > > > > > > > > > > >
> >>> > >> > > > > > > > > > > > > The issues that should be addressed is
> like
> >>> > >> > following:
> >>> > >> > > > > > > > > > > > >
> >>> > >> > > > > > > > > > > > > *1. Backward Compatibility*
> >>> > >> > > > > > > > > > > > >
> >>> > >> > > > > > > > > > > > > To support old consumers, we need to take
> a
> >>> > >> strategy
> >>> > >> > to
> >>> > >> > > > > > handle
> >>> > >> > > > > > > > the
> >>> > >> > > > > > > > > > old
> >>> > >> > > > > > > > > > > > > consumers. Current candidates are:
> >>> > >> > > > > > > > > > > > >
> >>> > >> > > > > > > > > > > > > - Bump API version
> >>> > >> > > > > > > > > > > > > - Leave unchanged: let the old clients
> fail.
> >>> > >> > > > > > > > > > > > > - Improve the error messages:
> >>> > >> > > > > > > > > > > > >
> >>> > >> > > > > > > > > > > > > *2. Dictionary Support*
> >>> > >> > > > > > > > > > > > >
> >>> > >> > > > > > > > > > > > > To support zstd's dictionary feature in
> the
> >>> > future
> >>> > >> > (if
> >>> > >> > > > > > needed),
> >>> > >> > > > > > > > we
> >>> > >> > > > > > > > > > need
> >>> > >> > > > > > > > > > > > to
> >>> > >> > > > > > > > > > > > > sketch how it should be and leave some
> room
> >>> for
> >>> > >> it.
> >>> > >> > As
> >>> > >> > > of
> >>> > >> > > > > > now,
> >>> > >> > > > > > > > > there
> >>> > >> > > > > > > > > > > has
> >>> > >> > > > > > > > > > > > > been no discussion on this topic yet.
> >>> > >> > > > > > > > > > > > >
> >>> > >> > > > > > > > > > > > > *3. License*
> >>> > >> > > > > > > > > > > > >
> >>> > >> > > > > > > > > > > > > To use this feature, we need to add
> license
> >>> of
> >>> > >> zstd
> >>> > >> > and
> >>> > >> > > > > > > zstd-jni
> >>> > >> > > > > > > > to
> >>> > >> > > > > > > > > > the
> >>> > >> > > > > > > > > > > > > project. (Thanks to Viktor Somogyi for
> >>> raising
> >>> > >> this
> >>> > >> > > > issue!)
> >>> > >> > > > > > It
> >>> > >> > > > > > > > > seems
> >>> > >> > > > > > > > > > > like
> >>> > >> > > > > > > > > > > > > what Apache Spark did would be a good
> >>> example
> >>> > but
> >>> > >> > there
> >>> > >> > > > has
> >>> > >> > > > > > > been
> >>> > >> > > > > > > > no
> >>> > >> > > > > > > > > > > > > discussion yet.
> >>> > >> > > > > > > > > > > > >
> >>> > >> > > > > > > > > > > > > You can find the details of the above
> >>> issues in
> >>> > >> the
> >>> > >> > KIP
> >>> > >> > > > > > > document.
> >>> > >> > > > > > > > > > > Please
> >>> > >> > > > > > > > > > > > > have a look when you are free, and give me
> >>> > >> feedback.
> >>> > >> > > All
> >>> > >> > > > > > kinds
> >>> > >> > > > > > > of
> >>> > >> > > > > > > > > > > > > participating are welcome.
> >>> > >> > > > > > > > > > > > >
> >>> > >> > > > > > > > > > > > > Best,
> >>> > >> > > > > > > > > > > > > Dongjin
> >>> > >> > > > > > > > > > > > >
> >>> > >> > > > > > > > > > > > > [^1]: At the time of writing, commit
> >>> 6b4fb8152.
> >>> > >> > > > > > > > > > > > >
> >>> > >> > > > > > > > > > > > > On Sat, Jul 14, 2018 at 10:45 PM Dongjin
> >>> Lee <
> >>> > >> > > > > > > dongjin@apache.org
> >>> > >> > > > > > > > >
> >>> > >> > > > > > > > > > > wrote:
> >>> > >> > > > > > > > > > > > >
> >>> > >> > > > > > > > > > > > > > Sorry for the late reply.
> >>> > >> > > > > > > > > > > > > >
> >>> > >> > > > > > > > > > > > > > In short, I could not submit the updated
> >>> KIP
> >>> > by
> >>> > >> the
> >>> > >> > > > > feature
> >>> > >> > > > > > > > > freeze
> >>> > >> > > > > > > > > > > > > > deadline of 2.0.0. For this reason, it
> >>> will
> >>> > not
> >>> > >> be
> >>> > >> > > > > included
> >>> > >> > > > > > > in
> >>> > >> > > > > > > > > the
> >>> > >> > > > > > > > > > > > 2.0.0
> >>> > >> > > > > > > > > > > > > > release and all discussion for this
> issue
> >>> were
> >>> > >> > > > postponed
> >>> > >> > > > > > > after
> >>> > >> > > > > > > > > the
> >>> > >> > > > > > > > > > > > release
> >>> > >> > > > > > > > > > > > > > of 2.0.0.
> >>> > >> > > > > > > > > > > > > >
> >>> > >> > > > > > > > > > > > > > I have been updating the PR following
> >>> recent
> >>> > >> > updates.
> >>> > >> > > > > Just
> >>> > >> > > > > > > > now, I
> >>> > >> > > > > > > > > > > > rebased
> >>> > >> > > > > > > > > > > > > > it against the latest trunk and updated
> >>> the
> >>> > zstd
> >>> > >> > > > version
> >>> > >> > > > > > into
> >>> > >> > > > > > > > > > 1.3.5.
> >>> > >> > > > > > > > > > > > If you
> >>> > >> > > > > > > > > > > > > > need some request, don't hesitate to
> >>> notify
> >>> > me.
> >>> > >> > (But
> >>> > >> > > > not
> >>> > >> > > > > > this
> >>> > >> > > > > > > > > > thread
> >>> > >> > > > > > > > > > > -
> >>> > >> > > > > > > > > > > > just
> >>> > >> > > > > > > > > > > > > > send me the message directly.)
> >>> > >> > > > > > > > > > > > > >
> >>> > >> > > > > > > > > > > > > > Best,
> >>> > >> > > > > > > > > > > > > > Dongjin
> >>> > >> > > > > > > > > > > > > >
> >>> > >> > > > > > > > > > > > > > On Tue, Jul 10, 2018 at 11:57 PM Bobby
> >>> Evans <
> >>> > >> > > > > > > bobby@apache.org
> >>> > >> > > > > > > > >
> >>> > >> > > > > > > > > > > wrote:
> >>> > >> > > > > > > > > > > > > >
> >>> > >> > > > > > > > > > > > > >> I there any update on this.  The
> >>> performance
> >>> > >> > > > > improvements
> >>> > >> > > > > > > are
> >>> > >> > > > > > > > > > quite
> >>> > >> > > > > > > > > > > > > >> impressive and I really would like to
> >>> stop
> >>> > >> forking
> >>> > >> > > > kafka
> >>> > >> > > > > > > just
> >>> > >> > > > > > > > to
> >>> > >> > > > > > > > > > get
> >>> > >> > > > > > > > > > > > this
> >>> > >> > > > > > > > > > > > > >> in.
> >>> > >> > > > > > > > > > > > > >>
> >>> > >> > > > > > > > > > > > > >> Thanks,
> >>> > >> > > > > > > > > > > > > >>
> >>> > >> > > > > > > > > > > > > >> Bobby
> >>> > >> > > > > > > > > > > > > >>
> >>> > >> > > > > > > > > > > > > >> On Wed, Jun 13, 2018 at 8:56 PM Dongjin
> >>> Lee <
> >>> > >> > > > > > > > dongjin@apache.org
> >>> > >> > > > > > > > > >
> >>> > >> > > > > > > > > > > > wrote:
> >>> > >> > > > > > > > > > > > > >>
> >>> > >> > > > > > > > > > > > > >> > Ismael,
> >>> > >> > > > > > > > > > > > > >> >
> >>> > >> > > > > > > > > > > > > >> > Oh, I forgot all of you are on
> working
> >>> > frenzy
> >>> > >> > for
> >>> > >> > > > 2.0!
> >>> > >> > > > > > No
> >>> > >> > > > > > > > > > problem,
> >>> > >> > > > > > > > > > > > take
> >>> > >> > > > > > > > > > > > > >> > your time. I am also working at
> another
> >>> > issue
> >>> > >> > now.
> >>> > >> > > > > Thank
> >>> > >> > > > > > > you
> >>> > >> > > > > > > > > for
> >>> > >> > > > > > > > > > > > > >> letting me
> >>> > >> > > > > > > > > > > > > >> > know.
> >>> > >> > > > > > > > > > > > > >> >
> >>> > >> > > > > > > > > > > > > >> > Best,
> >>> > >> > > > > > > > > > > > > >> > Dongjin
> >>> > >> > > > > > > > > > > > > >> >
> >>> > >> > > > > > > > > > > > > >> > On Wed, Jun 13, 2018, 11:44 PM Ismael
> >>> Juma
> >>> > <
> >>> > >> > > > > > > > ismaelj@gmail.com
> >>> > >> > > > > > > > > >
> >>> > >> > > > > > > > > > > > wrote:
> >>> > >> > > > > > > > > > > > > >> >
> >>> > >> > > > > > > > > > > > > >> > > Sorry for the delay Dongjin.
> >>> Everyone is
> >>> > >> busy
> >>> > >> > > > > > finalising
> >>> > >> > > > > > > > > > 2.0.0.
> >>> > >> > > > > > > > > > > > This
> >>> > >> > > > > > > > > > > > > >> KIP
> >>> > >> > > > > > > > > > > > > >> > > seems like a great candidate for
> >>> 2.1.0
> >>> > and
> >>> > >> > > > hopefully
> >>> > >> > > > > > > there
> >>> > >> > > > > > > > > > will
> >>> > >> > > > > > > > > > > be
> >>> > >> > > > > > > > > > > > > >> more
> >>> > >> > > > > > > > > > > > > >> > of
> >>> > >> > > > > > > > > > > > > >> > > a discussion next week. :)
> >>> > >> > > > > > > > > > > > > >> > >
> >>> > >> > > > > > > > > > > > > >> > > Ismael
> >>> > >> > > > > > > > > > > > > >> > >
> >>> > >> > > > > > > > > > > > > >> > > On Wed, 13 Jun 2018, 05:17 Dongjin
> >>> Lee, <
> >>> > >> > > > > > > > dongjin@apache.org
> >>> > >> > > > > > > > > >
> >>> > >> > > > > > > > > > > > wrote:
> >>> > >> > > > > > > > > > > > > >> > >
> >>> > >> > > > > > > > > > > > > >> > > > Hello. I just updated my draft
> >>> > >> > implementation:
> >>> > >> > > > > > > > > > > > > >> > > >
> >>> > >> > > > > > > > > > > > > >> > > > 1. Rebased to latest trunk
> (commit
> >>> > >> 5145d6b)
> >>> > >> > > > > > > > > > > > > >> > > > 2. Apply ZStd 1.3.4
> >>> > >> > > > > > > > > > > > > >> > > >
> >>> > >> > > > > > > > > > > > > >> > > > You can check out the
> >>> implementation
> >>> > from
> >>> > >> > here
> >>> > >> > > > > > > > > > > > > >> > > > <
> >>> > >> https://github.com/apache/kafka/pull/2267
> >>> > >> > >.
> >>> > >> > > If
> >>> > >> > > > > you
> >>> > >> > > > > > > > > > > experience
> >>> > >> > > > > > > > > > > > any
> >>> > >> > > > > > > > > > > > > >> > > problem
> >>> > >> > > > > > > > > > > > > >> > > > running it, don't hesitate to
> give
> >>> me a
> >>> > >> > > mention.
> >>> > >> > > > > > > > > > > > > >> > > >
> >>> > >> > > > > > > > > > > > > >> > > > Best,
> >>> > >> > > > > > > > > > > > > >> > > > Dongjin
> >>> > >> > > > > > > > > > > > > >> > > >
> >>> > >> > > > > > > > > > > > > >> > > > On Tue, Jun 12, 2018 at 6:50 PM
> >>> Dongjin
> >>> > >> Lee
> >>> > >> > <
> >>> > >> > > > > > > > > > > dongjin@apache.org
> >>> > >> > > > > > > > > > > > >
> >>> > >> > > > > > > > > > > > > >> > wrote:
> >>> > >> > > > > > > > > > > > > >> > > >
> >>> > >> > > > > > > > > > > > > >> > > > > Here is the short conclusion
> >>> about
> >>> > the
> >>> > >> > > license
> >>> > >> > > > > > > > problem:
> >>> > >> > > > > > > > > > *We
> >>> > >> > > > > > > > > > > > can
> >>> > >> > > > > > > > > > > > > >> use
> >>> > >> > > > > > > > > > > > > >> > > zstd
> >>> > >> > > > > > > > > > > > > >> > > > > and zstd-jni without any
> >>> problem, but
> >>> > >> we
> >>> > >> > > need
> >>> > >> > > > to
> >>> > >> > > > > > > > include
> >>> > >> > > > > > > > > > > their
> >>> > >> > > > > > > > > > > > > >> > license,
> >>> > >> > > > > > > > > > > > > >> > > > > e.g., BSD license.*
> >>> > >> > > > > > > > > > > > > >> > > > >
> >>> > >> > > > > > > > > > > > > >> > > > > Both of BSD 2 Clause License &
> 3
> >>> > Clause
> >>> > >> > > > License
> >>> > >> > > > > > > > requires
> >>> > >> > > > > > > > > > to
> >>> > >> > > > > > > > > > > > > >> include
> >>> > >> > > > > > > > > > > > > >> > the
> >>> > >> > > > > > > > > > > > > >> > > > > license used, and BSD 3 Clause
> >>> > License
> >>> > >> > > > requires
> >>> > >> > > > > > that
> >>> > >> > > > > > > > the
> >>> > >> > > > > > > > > > > name
> >>> > >> > > > > > > > > > > > of
> >>> > >> > > > > > > > > > > > > >> the
> >>> > >> > > > > > > > > > > > > >> > > > > contributor can't be used to
> >>> endorse
> >>> > or
> >>> > >> > > > promote
> >>> > >> > > > > > the
> >>> > >> > > > > > > > > > product.
> >>> > >> > > > > > > > > > > > > >> That's
> >>> > >> > > > > > > > > > > > > >> > it
> >>> > >> > > > > > > > > > > > > >> > > > > <
> >>> > >> > > > > > > > > > > > > >> > > >
> >>> > >> > > > > > > > > > > > > >> > >
> >>> > >> > > > > > > > > > > > > >> >
> >>> > >> > > > > > > > > > > > > >>
> >>> > >> > http://www.mikestratton.net/2011/12/is-bsd-license-
> >>> > >> > > > > > > > > > > > compatible-with-apache-2-0-license/
> >>> > >> > > > > > > > > > > > > >> > > > >
> >>> > >> > > > > > > > > > > > > >> > > > > - They are not listed in the
> >>> list of
> >>> > >> > > > prohibited
> >>> > >> > > > > > > > licenses
> >>> > >> > > > > > > > > > > > > >> > > > > <https://www.apache.org/legal/
> >>> > >> > > > > > > > resolved.html#category-x>
> >>> > >> > > > > > > > > > > also.
> >>> > >> > > > > > > > > > > > > >> > > > >
> >>> > >> > > > > > > > > > > > > >> > > > > Here is how Spark did for it
> >>> > >> > > > > > > > > > > > > >> > > > > <https://issues.apache.org/
> >>> > >> > > > > > jira/browse/SPARK-19112
> >>> > >> > > > > > > >:
> >>> > >> > > > > > > > > > > > > >> > > > >
> >>> > >> > > > > > > > > > > > > >> > > > > - They made a directory
> >>> dedicated to
> >>> > >> the
> >>> > >> > > > > > dependency
> >>> > >> > > > > > > > > > license
> >>> > >> > > > > > > > > > > > files
> >>> > >> > > > > > > > > > > > > >> > > > > <
> >>> > >> > > > > > > https://github.com/apache/spark/tree/master/licenses
> >>> > >> > > > > > > > >
> >>> > >> > > > > > > > > > and
> >>> > >> > > > > > > > > > > > added
> >>> > >> > > > > > > > > > > > > >> > > > licenses
> >>> > >> > > > > > > > > > > > > >> > > > > for Zstd
> >>> > >> > > > > > > > > > > > > >> > > > > <
> >>> > >> > > > > > > > > > > > > >> >
> >>> > >> > > > https://github.com/apache/spark/blob/master/licenses/
> >>> > >> > > > > > > > > > > > LICENSE-zstd.txt
> >>> > >> > > > > > > > > > > > > >> > > >
> >>> > >> > > > > > > > > > > > > >> > > > &
> >>> > >> > > > > > > > > > > > > >> > > > > Zstd-jni
> >>> > >> > > > > > > > > > > > > >> > > > > <
> >>> > >> > > > > > > > > > > > > >> > > >
> >>> > >> > > > > > > > > > > > > >> > >
> >>> > >> > > > > > > > > > > > > >> >
> >>> > >> > > > > > > > > > > > > >> https://github.com/apache/
> >>> > >> > > spark/blob/master/licenses/
> >>> > >> > > > > > > > > > > > LICENSE-zstd-jni.txt
> >>> > >> > > > > > > > > > > > > >> >
> >>> > >> > > > > > > > > > > > > >> > > > > .
> >>> > >> > > > > > > > > > > > > >> > > > > - Added a link to the original
> >>> > license
> >>> > >> > files
> >>> > >> > > > in
> >>> > >> > > > > > > > LICENSE.
> >>> > >> > > > > > > > > > > > > >> > > > > <
> >>> > >> > > > > https://github.com/apache/spark/pull/18805/files
> >>> > >> > > > > > >
> >>> > >> > > > > > > > > > > > > >> > > > >
> >>> > >> > > > > > > > > > > > > >> > > > > If needed, I can make a similar
> >>> > update.
> >>> > >> > > > > > > > > > > > > >> > > > >
> >>> > >> > > > > > > > > > > > > >> > > > > Thanks for pointing out this
> >>> problem,
> >>> > >> > > Viktor!
> >>> > >> > > > > Nice
> >>> > >> > > > > > > > > catch!
> >>> > >> > > > > > > > > > > > > >> > > > >
> >>> > >> > > > > > > > > > > > > >> > > > > Best,
> >>> > >> > > > > > > > > > > > > >> > > > > Dongjin
> >>> > >> > > > > > > > > > > > > >> > > > >
> >>> > >> > > > > > > > > > > > > >> > > > >
> >>> > >> > > > > > > > > > > > > >> > > > >
> >>> > >> > > > > > > > > > > > > >> > > > > On Mon, Jun 11, 2018 at 11:50
> PM
> >>> > >> Dongjin
> >>> > >> > > Lee <
> >>> > >> > > > > > > > > > > > dongjin@apache.org>
> >>> > >> > > > > > > > > > > > > >> > > wrote:
> >>> > >> > > > > > > > > > > > > >> > > > >
> >>> > >> > > > > > > > > > > > > >> > > > >> I greatly appreciate your
> >>> > >> comprehensive
> >>> > >> > > > > > reasoning.
> >>> > >> > > > > > > > so:
> >>> > >> > > > > > > > > +1
> >>> > >> > > > > > > > > > > > for b
> >>> > >> > > > > > > > > > > > > >> > until
> >>> > >> > > > > > > > > > > > > >> > > > now.
> >>> >