-
Type:
Bug
-
Resolution: Invalid
-
Priority:
Medium
-
None
-
Affects Version/s: 2.4.4
-
Component/s: None
-
Environment:
Apache + Tomcat + Postgres + RHEL4 + JDK1.5.0
Our confluence showed slow down recently. apache had large numbers of thread sitting in a writing state with no update being sent to the client.
We did a thread dump and noticed large numbers of the following messages:
...
"TP-Processor274" daemon prio=1 tid=0x0000002ba288e1b0 nid=0x506b waiting for monitor entry [0x0000000054d9a000..0x0000000054d9dd30]
at EDU.oswego.cs.dl.util.concurrent.BoundedLinkedQueue.put(BoundedLinkedQueue.java:298)
- waiting to lock <0x0000002aa64d8018> (a java.lang.Object)
at EDU.oswego.cs.dl.util.concurrent.QueuedExecutor.execute(QueuedExecutor.java:158)
at com.atlassian.confluence.ext.usage.event.UsageEventListener.addTask(UsageEventListener.java:63)
at com.atlassian.confluence.ext.usage.event.UsageEventListener.handleEvent(UsageEventListener.java:51)
...
"TP-Processor222" daemon prio=1 tid=0x0000002b9fb2a820 nid=0x4a9b waiting for monitor entry [0x0000000051c69000..0x0000000051c6cbb0]
at EDU.oswego.cs.dl.util.concurrent.BoundedLinkedQueue.put(BoundedLinkedQueue.java:298)
- waiting to lock <0x0000002aa64d8018> (a java.lang.Object)
at EDU.oswego.cs.dl.util.concurrent.QueuedExecutor.execute(QueuedExecutor.java:158)
at com.atlassian.confluence.ext.usage.event.UsageEventListener.addTask(UsageEventListener.java:63)
at com.atlassian.confluence.ext.usage.event.UsageEventListener.handleEvent(UsageEventListener.java:51)
...
"TP-Processor221" daemon prio=1 tid=0x0000002ba1844f30 nid=0x4a9a waiting for monitor entry [0x0000000051b68000..0x0000000051b6bb30]
at EDU.oswego.cs.dl.util.concurrent.BoundedLinkedQueue.put(BoundedLinkedQueue.java:298)
- waiting to lock <0x0000002aa64d8018> (a java.lang.Object)
at EDU.oswego.cs.dl.util.concurrent.QueuedExecutor.execute(QueuedExecutor.java:158)
at com.atlassian.confluence.ext.usage.event.UsageEventListener.addTask(UsageEventListener.java:63)
at com.atlassian.confluence.ext.usage.event.UsageEventListener.handleEvent(UsageEventListener.java:51)
...
Recurrance of UsageEventListener class leads us to Usage Tracking Plugin.
We noticed our Usage Tracking Plugin is shown as "Outdated".
Installed Version: 0.5
Latest Version: 0.7.2
Once we disabled the plugin. Confluence performance started getting back to normal.
It seems Outdated Usage Tracking Plugin caused locking issues.
