[
https://issues.apache.org/jira/browse/HDFS-4721?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13639252#comment-13639252
]
Varun Sharma commented on HDFS-4721:
------------------------------------
[~nkeywal]
>From what I think we may really have a scenario where we retry twice within
>the primary DN but I could be wrong and there could be some race conditions. I
>think the current behaviour is (from my look at the logs):
1) Try to recovery block from all data nodes
2) Hit a bad datanode - hit socket timeout (dfs.socket.timeout) which is
configurable.
3) Simply recover data block from the good datanodes and continue
Also, since primary DN is up and running, it always has the data block, so
there is at least one good datanode. I think optimizing recovery of > 1 block
within the primary DN could be in another JIRA. This one should be focused on
how to choose a) the primary datanode and b) the participating datanodes, at
the namenode.
Varun
> Speed up lease/block recovery when DN fails and a block goes into recovery
> --------------------------------------------------------------------------
>
> Key: HDFS-4721
> URL: https://issues.apache.org/jira/browse/HDFS-4721
> Project: Hadoop HDFS
> Issue Type: Improvement
> Components: namenode
> Affects Versions: 2.0.3-alpha
> Reporter: Varun Sharma
> Fix For: 2.0.4-alpha
>
> Attachments: 4721-hadoop2.patch, 4721-v2.patch, 4721-v3.patch,
> 4721-v4.patch, 4721-v5.patch
>
>
> This was observed while doing HBase WAL recovery. HBase uses append to write
> to its write ahead log. So initially the pipeline is setup as
> DN1 --> DN2 --> DN3
> This WAL needs to be read when DN1 fails since it houses the HBase
> regionserver for the WAL.
> HBase first recovers the lease on the WAL file. During recovery, we choose
> DN1 as the primary DN to do the recovery even though DN1 has failed and is
> not heartbeating any more.
> Avoiding the stale DN1 would speed up recovery and reduce hbase MTTR. There
> are two options.
> a) Ride on HDFS 3703 and if stale node detection is turned on, we do not
> choose stale datanodes (typically not heart beated for 20-30 seconds) as
> primary DN(s)
> b) We sort the replicas in order of last heart beat and always pick the ones
> which gave the most recent heart beat
> Going to the dead datanode increases lease + block recovery since the block
> goes into UNDER_RECOVERY state even though no one is recovering it actively.
> Please let me know if this makes sense. If yes, whether we should move
> forward with a) or b).
> Thanks
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
