[jira] [Comment Edited] (HBASE-16499) slow replication for small HBase clusters

["Ashish Singhi (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/HBASE-16499?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16426508#comment-16426508 ] 

Ashish Singhi edited comment on HBASE-16499 at 4/5/18 5:52 AM:
---------------------------------------------------------------

Thanks for the review. I have pushed the addendum to only master branch.

Addendum didn't apply to branch-2 and branch-2.0, as we have not committed HBASE-20273 in branch-2 and branch-2.0


was (Author: ashish singhi):
Thanks for the review. I have pushed the addendum to only master branch.

Addendum didn't apply as we have not committed HBASE-20273 in branch-2 and branch-2.0

> slow replication for small HBase clusters
> -----------------------------------------
>
>                 Key: HBASE-16499
>                 URL: https://issues.apache.org/jira/browse/HBASE-16499
>             Project: HBase
>          Issue Type: Bug
>          Components: Replication
>            Reporter: Vikas Vishwakarma
>            Assignee: Ashish Singhi
>            Priority: Critical
>             Fix For: 2.0.0
>
>         Attachments: HBASE-16499-addendum.patch, HBASE-16499.patch, HBASE-16499.patch
>
>
> For small clusters 10-20 nodes we recently observed that replication is progressing very slowly when we do bulk writes and there is lot of lag accumulation on AgeOfLastShipped / SizeOfLogQueue. From the logs we observed that the number of threads used for shipping wal edits in parallel comes from the following equation in HBaseInterClusterReplicationEndpoint
> int n = Math.min(Math.min(this.maxThreads, entries.size()/100+1),
>       replicationSinkMgr.getSinks().size());
> ... 
>       for (int i=0; i<n; i++) {
>         entryLists.add(new ArrayList<HLog.Entry>(entries.size()/n+1)); <-- batch size
>       }
> ...
>         for (int i=0; i<entryLists.size(); i++) {
>          .....
>             // RuntimeExceptions encountered here bubble up and are handled in ReplicationSource
>             pool.submit(createReplicator(entryLists.get(i), i));  <-- concurrency 
>             futures++;
>           }
>         }
> maxThreads is fixed & configurable and since we are taking min of the three values n gets decided based replicationSinkMgr.getSinks().size() when we have enough edits to replicate
> replicationSinkMgr.getSinks().size() is decided based on 
> int numSinks = (int) Math.ceil(slaveAddresses.size() * ratio);
> where ratio is this.ratio = conf.getFloat("replication.source.ratio", DEFAULT_REPLICATION_SOURCE_RATIO);
> Currently DEFAULT_REPLICATION_SOURCE_RATIO is set to 10% so for small clusters of size 10-20 RegionServers  the value we get for numSinks and hence n is very small like 1 or 2. This substantially reduces the pool concurrency used for shipping wal edits in parallel effectively slowing down replication for small clusters and causing lot of lag accumulation in AgeOfLastShipped. Sometimes it takes tens of hours to clear off the entire replication queue even after the client has finished writing on the source side. 
> We are running tests by varying replication.source.ratio and have seen multi-fold improvement in total replication time (will update the results here). I wanted to propose here that we should increase the default value for replication.source.ratio also so that we have sufficient concurrency even for small clusters. We figured it out after lot of iterations and debugging so probably slightly higher default will save the trouble. 



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)