You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@hbase.apache.org by "Guanghao Zhang (Jira)" <ji...@apache.org> on 2020/08/19 06:05:00 UTC
[jira] [Updated] (HBASE-24897) RegionReplicaFlushHandler should
handle NoServerForRegionException to avoid abort RegionServer
[ https://issues.apache.org/jira/browse/HBASE-24897?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Guanghao Zhang updated HBASE-24897:
-----------------------------------
Summary: RegionReplicaFlushHandler should handle NoServerForRegionException to avoid abort RegionServer (was: Fix flaky test TestRegionReplicaReplicationEndpoint)
> RegionReplicaFlushHandler should handle NoServerForRegionException to avoid abort RegionServer
> ----------------------------------------------------------------------------------------------
>
> Key: HBASE-24897
> URL: https://issues.apache.org/jira/browse/HBASE-24897
> Project: HBase
> Issue Type: Bug
> Reporter: Guanghao Zhang
> Assignee: Guanghao Zhang
> Priority: Major
>
> Debug flaky test TestRegionReplicaReplicationEndpoint, I found the RS aborted because RegionReplicaFlushHandler flush failed. When create a new table with region replica, the assign order may be:
> # assign 0002 replica region and trigger primary region flush.
> # assign 0001 replica region and trigger primary region flush.
> # assign primary region.
> But the primary region flush may failed because the primary region not opened now. So it may abort the RS......
>
> {code:java}
> 2020-08-18 16:56:30,041 INFO [RS_OPEN_REGION-regionserver/hao-OptiPlex-7050:0-0] handler.AssignRegionHandler(141): Opened testRegionReplicaReplicationIgnoresDisabledTables_drop_false_disabledReplication_false,,1597740978463_0002.66e9757a05fbae7623cfea3369fc8354.
> 2020-08-18 16:56:30,558 INFO [RS_OPEN_REGION-regionserver/hao-OptiPlex-7050:0-0] handler.AssignRegionHandler(141): Opened testRegionReplicaReplicationIgnoresDisabledTables_drop_false_disabledReplication_false,,1597740978463_0001.22ff45423b0f1f0e93794f673449d140.
> 2020-08-18 16:56:31,192 INFO [RS_OPEN_REGION-regionserver/hao-OptiPlex-7050:0-0] handler.AssignRegionHandler(141): Opened testRegionReplicaReplicationIgnoresDisabledTables_drop_false_disabledReplication_false,,1597740978463.901f9cd06bbf27ef7c2d70b5af725cd2.
> 2020-08-18 16:58:53,857 ERROR [RS_REGION_REPLICA_FLUSH_OPS-regionserver/hao-OptiPlex-7050:0-0] helpers.MarkerIgnoringBase(159): ***** ABORTING region server hao-optiplex-7050,36368,1597740961432: ServerAborting because an exception was thrown *****
> org.apache.hadoop.hbase.client.NoServerForRegionException: No server address listed in hbase:meta for region testRegionReplicaReplicationWithReplicas_10,,1597741128945.0f541dc1a7ca64797c4cf054adb9edfb. containing row
> at org.apache.hadoop.hbase.client.ConnectionImplementation.locateRegionInMeta(ConnectionImplementation.java:926)
> at org.apache.hadoop.hbase.client.ConnectionImplementation.locateRegion(ConnectionImplementation.java:784)
> at org.apache.hadoop.hbase.client.ConnectionUtils$ShortCircuitingClusterConnection.locateRegion(ConnectionUtils.java:140)
> at org.apache.hadoop.hbase.client.RegionAdminServiceCallable.getRegionLocations(RegionAdminServiceCallable.java:147)
> at org.apache.hadoop.hbase.client.RegionAdminServiceCallable.getLocation(RegionAdminServiceCallable.java:98)
> at org.apache.hadoop.hbase.client.RegionAdminServiceCallable.prepare(RegionAdminServiceCallable.java:84)
> at org.apache.hadoop.hbase.client.FlushRegionCallable.prepare(FlushRegionCallable.java:62)
> at org.apache.hadoop.hbase.client.RpcRetryingCallerImpl.callWithRetries(RpcRetryingCallerImpl.java:105)
> at org.apache.hadoop.hbase.regionserver.handler.RegionReplicaFlushHandler.triggerFlushInPrimaryRegion(RegionReplicaFlushHandler.java:129)
> at org.apache.hadoop.hbase.regionserver.handler.RegionReplicaFlushHandler.process(RegionReplicaFlushHandler.java:78)
> at org.apache.hadoop.hbase.executor.EventHandler.run(EventHandler.java:104)
> at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
> at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
> at java.lang.Thread.run(Thread.java:748)
> {code}
> I thought the fix should be assign primary region firstly when enable region replica featue. Will check the implmenation of region replica.
>
--
This message was sent by Atlassian Jira
(v8.3.4#803005)