[
https://issues.apache.org/jira/browse/HBASE-11096?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13985069#comment-13985069
]
Qiang Tian commented on HBASE-11096:
------------------------------------
Hi [~apurtell],
Thanks for response.
Yes, I will. the fix is simple.. this is part of work for HBASE-10289,
during my test I just find another small bug--the HRegionServer.stop is called
twice, the fix looks simple as well, but I better study the related code
first..:-)
the first time:
HRegionServer.run:
} else if (this.stopping) {
boolean allUserRegionsOffline = areAllUserRegionsOffline();
if (allUserRegionsOffline) {
// Set stopped if no requests since last time we went around the
loop.
// The remaining meta regions will be closed on our way out.
if (oldRequestCount == this.requestCount.get()) {
stop("Stopped; only catalog regions remaining online");
break;
}
the second time in shutdown hook:
at
org.apache.hadoop.hbase.coprocessor.CoprocessorHost.shutdown(CoprocessorHost.java:273)
at
org.apache.hadoop.hbase.regionserver.RegionServerCoprocessorHost.preStop(RegionServerCoproces
sorHost.java:63)
at
org.apache.hadoop.hbase.regionserver.HRegionServer.stop(HRegionServer.java:1680)
at
org.apache.hadoop.hbase.regionserver.ShutdownHook$ShutdownHookThread.run(ShutdownHook.java:114)
at
org.apache.hadoop.util.ShutdownHookManager$1.run(ShutdownHookManager.java:54)
2014-04-29 18:56:02,617 WARN
org.apache.hadoop.hbase.coprocessor.CoprocessorHost: Not stopping coprocess
or org.apache.hadoop.hbase.jmxAgent because not active (state=STOPPED)
2014-04-29 18:56:02,617 INFO
org.apache.hadoop.hbase.regionserver.HRegionServer: STOPPED: Shutdown hook
2014-04-29 18:56:02,617 INFO org.apache.hadoop.hbase.regionserver.ShutdownHook:
Starting fs shutdown hoo
k thread.
> stop method of Master and RegionServer coprocessor is not invoked
> ------------------------------------------------------------------
>
> Key: HBASE-11096
> URL: https://issues.apache.org/jira/browse/HBASE-11096
> Project: HBase
> Issue Type: Bug
> Affects Versions: 0.96.2, 0.98.1, 0.94.19
> Reporter: Qiang Tian
> Assignee: Qiang Tian
> Fix For: 0.99.0, 0.96.3, 0.94.20, 0.98.3
>
>
> the stop method of coprocessor specified by
> "hbase.coprocessor.master.classes" and
> "hbase.coprocessor.regionserver.classes" is not invoked.
> If coprocessor allocates OS resources, it could cause master/regionserver
> resource leak or hang during exit.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
