You are viewing a plain text version of this content. The canonical link for it is here.
Posted to hdfs-user@hadoop.apache.org by Lukas Kairies <lu...@googlemail.com> on 2013/08/27 16:58:26 UTC
client side I/O buffering
Hello,
I know that the client caches write requests before they send to a
datanode and that the client uses read-ahead, but where exactly is this
implemented, Hadoop or HDFS? Or better question, will write cache and
read-ahead also available when Hadoop uses another filesysten than HDFS?
Thanks,
Lukas K.
Re: client side I/O buffering
Posted by Arpit Agarwal <aa...@hortonworks.com>.
Hi Lukas,
HDFS read-ahead/write caching is provided by DFSInputStream and
DFSOutputStream respectively (see CachingStrategy).
Other file system providers need to implement their own caching strategy.
FileSystem#open and FileContext#open just return a seekable input stream
with no read-ahead assumptions.
-Arpit
On Tue, Aug 27, 2013 at 7:58 AM, Lukas Kairies <
lukas.xtreemfs@googlemail.com> wrote:
> Hello,
>
> I know that the client caches write requests before they send to a
> datanode and that the client uses read-ahead, but where exactly is this
> implemented, Hadoop or HDFS? Or better question, will write cache and
> read-ahead also available when Hadoop uses another filesysten than HDFS?
>
> Thanks,
> Lukas K.
>
>
>
--
CONFIDENTIALITY NOTICE
NOTICE: This message is intended for the use of the individual or entity to
which it is addressed and may contain information that is confidential,
privileged and exempt from disclosure under applicable law. If the reader
of this message is not the intended recipient, you are hereby notified that
any printing, copying, dissemination, distribution, disclosure or
forwarding of this communication is strictly prohibited. If you have
received this communication in error, please contact the sender immediately
and delete it from your system. Thank You.
Re: client side I/O buffering
Posted by Arpit Agarwal <aa...@hortonworks.com>.
Hi Lukas,
HDFS read-ahead/write caching is provided by DFSInputStream and
DFSOutputStream respectively (see CachingStrategy).
Other file system providers need to implement their own caching strategy.
FileSystem#open and FileContext#open just return a seekable input stream
with no read-ahead assumptions.
-Arpit
On Tue, Aug 27, 2013 at 7:58 AM, Lukas Kairies <
lukas.xtreemfs@googlemail.com> wrote:
> Hello,
>
> I know that the client caches write requests before they send to a
> datanode and that the client uses read-ahead, but where exactly is this
> implemented, Hadoop or HDFS? Or better question, will write cache and
> read-ahead also available when Hadoop uses another filesysten than HDFS?
>
> Thanks,
> Lukas K.
>
>
>
--
CONFIDENTIALITY NOTICE
NOTICE: This message is intended for the use of the individual or entity to
which it is addressed and may contain information that is confidential,
privileged and exempt from disclosure under applicable law. If the reader
of this message is not the intended recipient, you are hereby notified that
any printing, copying, dissemination, distribution, disclosure or
forwarding of this communication is strictly prohibited. If you have
received this communication in error, please contact the sender immediately
and delete it from your system. Thank You.
Re: client side I/O buffering
Posted by Arpit Agarwal <aa...@hortonworks.com>.
Hi Lukas,
HDFS read-ahead/write caching is provided by DFSInputStream and
DFSOutputStream respectively (see CachingStrategy).
Other file system providers need to implement their own caching strategy.
FileSystem#open and FileContext#open just return a seekable input stream
with no read-ahead assumptions.
-Arpit
On Tue, Aug 27, 2013 at 7:58 AM, Lukas Kairies <
lukas.xtreemfs@googlemail.com> wrote:
> Hello,
>
> I know that the client caches write requests before they send to a
> datanode and that the client uses read-ahead, but where exactly is this
> implemented, Hadoop or HDFS? Or better question, will write cache and
> read-ahead also available when Hadoop uses another filesysten than HDFS?
>
> Thanks,
> Lukas K.
>
>
>
--
CONFIDENTIALITY NOTICE
NOTICE: This message is intended for the use of the individual or entity to
which it is addressed and may contain information that is confidential,
privileged and exempt from disclosure under applicable law. If the reader
of this message is not the intended recipient, you are hereby notified that
any printing, copying, dissemination, distribution, disclosure or
forwarding of this communication is strictly prohibited. If you have
received this communication in error, please contact the sender immediately
and delete it from your system. Thank You.
Re: client side I/O buffering
Posted by Arpit Agarwal <aa...@hortonworks.com>.
Hi Lukas,
HDFS read-ahead/write caching is provided by DFSInputStream and
DFSOutputStream respectively (see CachingStrategy).
Other file system providers need to implement their own caching strategy.
FileSystem#open and FileContext#open just return a seekable input stream
with no read-ahead assumptions.
-Arpit
On Tue, Aug 27, 2013 at 7:58 AM, Lukas Kairies <
lukas.xtreemfs@googlemail.com> wrote:
> Hello,
>
> I know that the client caches write requests before they send to a
> datanode and that the client uses read-ahead, but where exactly is this
> implemented, Hadoop or HDFS? Or better question, will write cache and
> read-ahead also available when Hadoop uses another filesysten than HDFS?
>
> Thanks,
> Lukas K.
>
>
>
--
CONFIDENTIALITY NOTICE
NOTICE: This message is intended for the use of the individual or entity to
which it is addressed and may contain information that is confidential,
privileged and exempt from disclosure under applicable law. If the reader
of this message is not the intended recipient, you are hereby notified that
any printing, copying, dissemination, distribution, disclosure or
forwarding of this communication is strictly prohibited. If you have
received this communication in error, please contact the sender immediately
and delete it from your system. Thank You.