[
https://issues.apache.org/jira/browse/HBASE-26812?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17509088#comment-17509088
]
Lars Hofhansl edited comment on HBASE-26812 at 3/18/22, 10:28 PM:
------------------------------------------------------------------
FYI... There's a problem with scanning too (I guess we knew). So PHOENIX-6501
is also currently broken with HBase 2.
BTW. In that case wrapping with a null call context did *not* fix all problems.
The context is used to resolve the right user for scanner.next(...) if that is
missing it gets the system user, which is incredibly slow.
The savings from this are questionable anyway. Even on a local machine with all
regions locally I hardly see a perf difference with disabling this.
I'd vote for ripping it out.
was (Author: lhofhansl):
FYI... There's a problem with scanning too (I guess we knew). So PHOENIX-6501
is also currently broken with HBase 2.
BTW. In that case wrapping with a null call context did *not* help. The context
is used to resolve the right user for scanner.next(...) if that is missing it
gets the system user, which is incredibly slow.
The savings from this are questionable anyway. Even on a local machine with all
regions locally I hardly see a perf difference with disabling this.
I'd vote for ripping it out.
> ShortCircuitingClusterConnection fails to close RegionScanners when making
> short-circuited calls
> ------------------------------------------------------------------------------------------------
>
> Key: HBASE-26812
> URL: https://issues.apache.org/jira/browse/HBASE-26812
> Project: HBase
> Issue Type: Bug
> Affects Versions: 2.4.9
> Reporter: Lars Hofhansl
> Priority: Critical
>
> Just ran into this on the Phoenix side.
> We retrieve a Connection via
> {{{}RegionCoprocessorEnvironment.createConnection... getTable(...){}}}. And
> then call get on that table. The Get's key happens to be local. Now each call
> to table.get() leaves an open StoreScanner around forever. (verified with a
> memory profiler).
> There references are held via
> RegionScannerImpl.storeHeap.scannersForDelayedClose. Eventially the
> RegionServer goes into a GC of death and can only ended with kill -9.
> The reason appears to be that in this case there is no currentCall context.
> Some time in 2.x the Rpc handler/call was made responsible for closing open
> region scanners, but we forgot to handle {{ShortCircuitingClusterConnection}}
> It's not immediately clear how to fix this. But it does make
> ShortCircuitingClusterConnection useless and dangerous. If you use it, you
> *will* create a giant memory leak.
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
