[jira] [Comment Edited] (IGNITE-3748) Data rebalancing of large cache can hang out.

["Anton Vinogradov (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/IGNITE-3748?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15650777#comment-15650777 ] 

Anton Vinogradov edited comment on IGNITE-3748 at 11/9/16 12:18 PM:
--------------------------------------------------------------------

Possible, problem at org.apache.ignite.internal.processors.cache.distributed.dht.preloader.GridDhtPreloader#evictPartitionAsync
{noformat}
if (part != null)
   part.tryEvict();
{noformat}
should be replaced with 
{noformat}
try {
   part.tryEvict();
}
catch (Throwable ex) {
   LT.error(log, ex, "Partition eviction failed, this can cause grid hang.");
}
{noformat}


was (Author: avinogradov):
Possible, problem at org.apache.ignite.internal.processors.cache.distributed.dht.preloader.GridDhtPreloader#evictPartitionAsync
{nofirmat}
if (part != null)
   part.tryEvict();
{noformat}
should be replaced with 
{noformat}
try {
   part.tryEvict();
}
catch (Throwable ex) {
   LT.error(log, ex, "Partition eviction failed, this can cause grid hang.");
}
{noformat}

> Data rebalancing of large cache can hang out.
> ---------------------------------------------
>
>                 Key: IGNITE-3748
>                 URL: https://issues.apache.org/jira/browse/IGNITE-3748
>             Project: Ignite
>          Issue Type: Bug
>            Reporter: Vladislav Pyatkov
>            Assignee: Anton Vinogradov
>         Attachments: Apache.config.xml, default-config.xml, logs.zip, thread_dump.zip
>
>
> After the node (_CO3SCH050520537_) disconnect from cluster and starts again, rebalacing partitions of cache works, until node will faill by timeout.
> In the log of node which joining:
> {noformat}
> [2016.08.17 03:28:52,754 PDT][INFO ][sys-#122%null%][GridDhtPartitionDemander] <cache_raw_gbievent> Completed (final) rebalancing [cache=cache_raw_gbievent, fromNode=14666739-08e3-4188-ace7-9e47940bed70, topology=AffinityTopologyVersion [topVer=205, minorTopVer=0], time=336336 ms]
> {noformat}
> In other server nodes:
> {noformat}
> [2016.08.17 03:29:04,413 PDT][WARN ][exchange-worker-#176%null%][GridDhtPartitionsExchangeFuture] Failed to wait for partition release future [topVer=AffinityTopologyVersion [topVer=205, minorTopVer=1], node=14666739-08e3-4188-ace7-9e47940bed70].
> {noformat}
> See [the thread for details | http://apache-ignite-users.70518.x6.nabble.com/Failed-to-wait-for-initial-partition-map-exchange-tt6252.html#a7171]



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)