[
https://issues.apache.org/jira/browse/HDFS-1783?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13283154#comment-13283154
]
Lars Hofhansl commented on HDFS-1783:
-------------------------------------
So the proposal here is to have the client parallelize the requests to the
replicas.
How about the option of having the first DN in the chain do that?
The client interacts with the DNs as before, but the first DN would parallelize
to all (or some) downstream DNs.
I.e. if there are 3 replicas the first DN would contact the 2nd and 3rd DN in
parallel. Looking at the PacketResponder code, that shouldn't be too hard to
add.
Would be less parallel than doing it all in the client, but the client would
not have to be aware of the changes. On the flipside having this all in the
client avoids DN changes.
As Dhruba points out, that would be quite useful for the HBase WALs, especially
with HDFS-744.
> Ability for HDFS client to write replicas in parallel
> -----------------------------------------------------
>
> Key: HDFS-1783
> URL: https://issues.apache.org/jira/browse/HDFS-1783
> Project: Hadoop HDFS
> Issue Type: Improvement
> Components: hdfs client
> Reporter: dhruba borthakur
> Assignee: dhruba borthakur
>
> The current implementation of HDFS pipelines the writes to the three
> replicas. This introduces some latency for realtime latency sensitive
> applications. An alternate implementation that allows the client to write all
> replicas in parallel gives much better response times to these applications.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
