You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@zookeeper.apache.org by Patrick Hunt <ph...@apache.org> on 2016/02/11 18:41:24 UTC
Interesting paper on crash recovery - incl ZK
I noticed this on "the morning paper" this morning:
http://blog.acolyer.org/2016/02/11/fs-not-equal/
https://www.usenix.org/system/files/conference/osdi14/osdi14-paper-pillai.pdf
Seems they found a couple of issues that ZK is vulnerable to that we
should consider...
Patrick
Re: Interesting paper on crash recovery - incl ZK
Posted by Chris Nauroth <cn...@hortonworks.com>.
Patrick's linked paper was different: "All File Systems Are Not Created
Equal: On the Complexity of Crafting Crash-Consistent Applications". This
one is "Simple testing can prevent most critical failures-- An analysis of
production failures in distributed data-intensive systems". This one is
interesting too though, so thanks for sharing again.
--Chris Nauroth
On 2/11/16, 2:44 PM, "Flavio Junqueira" <fp...@apache.org> wrote:
>I pointed to that work after attending the presentation back in 2014,
>here is the message I sent to the list back then:
>
>http://mail-archives.apache.org/mod_mbox/zookeeper-dev/201410.mbox/%3C1412
>720688.91993.YahooMailNeo@web142305.mail.bf1.yahoo.com%3E
><http://mail-archives.apache.org/mod_mbox/zookeeper-dev/201410.mbox/%3C141
>2720688.91993.YahooMailNeo@web142305.mail.bf1.yahoo.com%3E>
>
>-Flavio
>
>> On 11 Feb 2016, at 21:44, Chris Nauroth <cn...@hortonworks.com>
>>wrote:
>>
>> This is a good paper. I read it a while ago more in the context of HDFS
>> work. I forgot that they also analyzed ZooKeeper. Here is a link to
>> their USENIX talk for those who prefer to digest it in that form.
>>
>> https://www.youtube.com/watch?v=SVYegdh2CbE
>>
>>
>> --Chris Nauroth
>>
>>
>>
>>
>> On 2/11/16, 9:41 AM, "Patrick Hunt" <ph...@apache.org> wrote:
>>
>>> I noticed this on "the morning paper" this morning:
>>> http://blog.acolyer.org/2016/02/11/fs-not-equal/
>>>
>>>
>>>https://www.usenix.org/system/files/conference/osdi14/osdi14-paper-pilla
>>>i.
>>> pdf
>>>
>>> Seems they found a couple of issues that ZK is vulnerable to that we
>>> should consider...
>>>
>>> Patrick
>>>
>>
>
Re: Interesting paper on crash recovery - incl ZK
Posted by Flavio Junqueira <fp...@apache.org>.
I pointed to that work after attending the presentation back in 2014, here is the message I sent to the list back then:
http://mail-archives.apache.org/mod_mbox/zookeeper-dev/201410.mbox/%3C1412720688.91993.YahooMailNeo@web142305.mail.bf1.yahoo.com%3E <http://mail-archives.apache.org/mod_mbox/zookeeper-dev/201410.mbox/%3C1412720688.91993.YahooMailNeo@web142305.mail.bf1.yahoo.com%3E>
-Flavio
> On 11 Feb 2016, at 21:44, Chris Nauroth <cn...@hortonworks.com> wrote:
>
> This is a good paper. I read it a while ago more in the context of HDFS
> work. I forgot that they also analyzed ZooKeeper. Here is a link to
> their USENIX talk for those who prefer to digest it in that form.
>
> https://www.youtube.com/watch?v=SVYegdh2CbE
>
>
> --Chris Nauroth
>
>
>
>
> On 2/11/16, 9:41 AM, "Patrick Hunt" <ph...@apache.org> wrote:
>
>> I noticed this on "the morning paper" this morning:
>> http://blog.acolyer.org/2016/02/11/fs-not-equal/
>>
>> https://www.usenix.org/system/files/conference/osdi14/osdi14-paper-pillai.
>> pdf
>>
>> Seems they found a couple of issues that ZK is vulnerable to that we
>> should consider...
>>
>> Patrick
>>
>
Re: Interesting paper on crash recovery - incl ZK
Posted by Chris Nauroth <cn...@hortonworks.com>.
This is a good paper. I read it a while ago more in the context of HDFS
work. I forgot that they also analyzed ZooKeeper. Here is a link to
their USENIX talk for those who prefer to digest it in that form.
https://www.youtube.com/watch?v=SVYegdh2CbE
--Chris Nauroth
On 2/11/16, 9:41 AM, "Patrick Hunt" <ph...@apache.org> wrote:
>I noticed this on "the morning paper" this morning:
>http://blog.acolyer.org/2016/02/11/fs-not-equal/
>
>https://www.usenix.org/system/files/conference/osdi14/osdi14-paper-pillai.
>pdf
>
>Seems they found a couple of issues that ZK is vulnerable to that we
>should consider...
>
>Patrick
>