EHC ❯ Write-behind IllegalMonitorState when last client terminates
-
Bug
-
Status: Closed
-
1 Critical
-
Resolution: Fixed
-
ehcache-core
-
-
alexsnaps
-
Reporter: gkeim
-
April 30, 2011
-
0
-
Watchers: 2
-
July 27, 2012
-
September 27, 2011
Description
SEVERE: java.lang.IllegalMonitorStateException at java.util.concurrent.locks.ReentrantReadWriteLock$Sync.tryRelease(ReentrantReadWriteLock.java:317) at java.util.concurrent.locks.AbstractQueuedSynchronizer.release(AbstractQueuedSynchronizer.java:1239) at java.util.concurrent.locks.ReentrantReadWriteLock$WriteLock.__RWL__tc_unlock(ReentrantReadWriteLock.java:992) at java.util.concurrent.locks.ReentrantReadWriteLock$WriteLock.unlock(ReentrantReadWriteLock.java) at org.terracotta.locking.TerracottaReadWriteLock$WriteLock.unlock(TerracottaReadWriteLock.java:359) at org.terracotta.async.ProcessingBucketGroup.quarantineItemsFromDeadBuckets(ProcessingBucketGroup.java:257) at org.terracotta.async.ProcessingBucket.processItems(ProcessingBucket.java:466) at org.terracotta.async.ProcessingBucket.access$400(ProcessingBucket.java:37) at org.terracotta.async.ProcessingBucket$ProcessingThread.run(ProcessingBucket.java:176) at java.lang.Thread.run(Thread.java:662)
This only occurs when the last client shuts down. If no data is being lost, then I would just consider this noise. You can reproduce this with EhcachePounder. Start 3 nodes, turn up the speed, then after a while terminate each serially.
Comments
Fiona OShea 2011-05-03
Tim Eck 2011-05-16
just speculating really but this looks like more of a generic problem with the async stuff in the toolkit as opposed to something specific to ehcache
Alexander Snaps 2011-06-28
Yes, more on this here : DEV-5924
Fiona OShea 2011-09-26
Please also merge to ehcache-core-2.4.x and ehcache-core-2.4.6-RC
thanks
Alexander Snaps 2011-09-27
Merged to toolkit os tc-3.53 & tc-3.5.4 branches
Maybe a better error message?