[teiid-issues] [JBoss JIRA] (TEIID-3106) BufferManager Cleaner consuming 90% of total CPU time
Mark Addleman (JIRA)
issues at jboss.org
Tue Sep 2 18:41:59 EDT 2014
[ https://issues.jboss.org/browse/TEIID-3106?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12998148#comment-12998148 ]
Mark Addleman commented on TEIID-3106:
--------------------------------------
> But that would imply that more than 2 billion reads had occurred, which seems unlikely for start up.
Agreed. I was looking through invokers of new CacheKey to see if I could spot an underflow from them. I didn't but I did see a possible anomaly: The starting values for lastAccess and orderingValue are zero from almost all clients except org.teiid.common.buffer.impl.BufferFrontedFileStoreCache.get(PhysicalInfo, Long, WeakReference<? extends Serializer<?>>) where they are one. This strikes me as odd but there could be good reason for it.
> BufferManager Cleaner consuming 90% of total CPU time
> -----------------------------------------------------
>
> Key: TEIID-3106
> URL: https://issues.jboss.org/browse/TEIID-3106
> Project: Teiid
> Issue Type: Bug
> Security Level: Public(Everyone can see)
> Components: Query Engine
> Affects Versions: 8.7
> Environment: zOS
> Reporter: Devesh Mishra
> Assignee: Steven Hawkins
> Labels: teiid-engine
> Attachments: BufferManagerImpl.java, CacheKey_Loog.patch
>
>
> BufferManager Cleaner thread is consuming almost all of the CPU utilized by the jboss process. Thread dump shows following information.
> 3XMTHREADINFO "BufferManager Cleaner" J9VMThread:0x0000004C41CFEB00, j9thread_t:0x0000004C52B85AE0, java/lang/Thread:0x000000481C036E20, state:CW, prio=5
> 3XMJAVALTHREAD (java/lang/Thread getId:0x76, isDaemon:true)
> 3XMTHREADINFO1 (native thread ID:0x3AEC2600, native priority:0x5, native policy:UNKNOWN)
> 3XMHEAPALLOC Heap bytes allocated since last GC cycle=2609184 (0x27D020)
> 3XMTHREADINFO3 Java callstack:
> 4XESTACKTRACE at java/util/concurrent/ConcurrentSkipListMap.doRemove(ConcurrentSkipListMap.java:1070(Compiled Code))
> 4XESTACKTRACE at java/util/concurrent/ConcurrentSkipListMap.remove(ConcurrentSkipListMap.java:1659(Compiled Code))
> 4XESTACKTRACE at org/teiid/common/buffer/impl/LrfuEvictionQueue.remove(LrfuEvictionQueue.java:60(Compiled Code))
> 4XESTACKTRACE at org/teiid/common/buffer/impl/BufferManagerImpl.doEvictions(BufferManagerImpl.java:854(Compiled Code))
> 5XESTACKTRACE (entered lock: org/teiid/common/buffer/CacheEntry at 0x00000048393C2598, entry count: 1)
> 4XESTACKTRACE at org/teiid/common/buffer/impl/BufferManagerImpl$Cleaner.run(BufferManagerImpl.java:108)
> 4XESTACKTRACE at java/util/TimerThread.mainLoop(Timer.java:555)
> 4XESTACKTRACE at java/util/TimerThread.run(Timer.java:505)
> When we added log statements around the BufferManagerImpl.doEvictions() it loops through the remove and firstEntry loop.
> Forum discussion link : https://community.jboss.org/message/901792
--
This message was sent by Atlassian JIRA
(v6.3.1#6329)
More information about the teiid-issues
mailing list