[
https://issues.jboss.org/browse/TEIID-3070?page=com.atlassian.jira.plugin...
]
Kylin Soong edited comment on TEIID-3070 at 10/8/14 11:30 PM:
--------------------------------------------------------------
Hi Steven
I have talked with norman that seems netty 3.6.x have this issue(DV 6.0 use netty 3.6.6,
DV 6.1 use netty 3.6.7), not sure other versions.
Teiid 8.7.x use bom to contorl netty version, as Bugzilla updates, the issue being
handled in wildfly and EAP upstream, no changes are needed in teiid side.
Netty 3.6.10 has been
released(https://github.com/netty/netty/releases/tag/netty-3.6.10.Final), for current
master netty 8.9, I have tested update netty version to 3.6.10, but execute build failed,
do you think we need upgrate the netty version? if yes, I will do more investigation.
was (Author: kylin):
Hi Steven
I have talked with norman that seems netty 3.6.x have this issue(DV 6.0 use netty 3.6.6,
DV 6.1 use netty 3.6.7), not sure other versions.
Teiid 8.7.x use bom to contorl netty version, as Bugzilla updates, the issue being
handled in wildfly and EAP upstream, no changes are needed in teiid side.
Netty 3.6.10 has been
released(https://github.com/netty/netty/releases/tag/netty-3.6.10.Final), for current
master netty 8.9, I think we can update netty version to 3.6.10 directly.
Netty worker threads cause system high cpu
------------------------------------------
Key: TEIID-3070
URL:
https://issues.jboss.org/browse/TEIID-3070
Project: Teiid
Issue Type: Bug
Components: Server
Affects Versions: 8.7
Environment: JDV 6.1.0.DR2
JDV 6.0
Reporter: Kylin Soong
Assignee: Kylin Soong
I have hit this issue many times in deploying a VDB to JDV, All CPU be used by netty
worker threads:
~~~
"New I/O worker #3" daemon prio=10 tid=0x00007fc71c0d2000 nid=0x320b runnable
[0x00007fc7074f3000]
java.lang.Thread.State: RUNNABLE
at sun.nio.ch.EPollArrayWrapper.epollWait(Native Method)
at sun.nio.ch.EPollArrayWrapper.poll(EPollArrayWrapper.java:269)
at sun.nio.ch.EPollSelectorImpl.doSelect(EPollSelectorImpl.java:79)
at sun.nio.ch.SelectorImpl.lockAndDoSelect(SelectorImpl.java:87)
- locked <0x00000000c42b4318> (a sun.nio.ch.Util$2)
- locked <0x00000000c42b4328> (a java.util.Collections$UnmodifiableSet)
- locked <0x00000000c42b42d0> (a sun.nio.ch.EPollSelectorImpl)
at sun.nio.ch.SelectorImpl.select(SelectorImpl.java:98)
at org.jboss.netty.channel.socket.nio.SelectorUtil.select(SelectorUtil.java:64)
~~~
Seems this is a exist netty issue, search "jboss netty high cpu" via google we
can get lots of result.
Below link have detailed depiction about this issue;
https://github.com/kylinsoong/teiid-samples/tree/master/highcpu
--
This message was sent by Atlassian JIRA
(v6.3.1#6329)