Alfresco getting shutdown frequently

[anndrewtr]

Hi;

We are having 2 Alfresco servers which are running in cluster. One of the Alfresco server is getting shutdown frequently. Sometimes there is some log while getting shutdown but most of the times there is no log trace.

The last log message which was been printed on shutdown was:

Heap
par new generation   total 943744K, used 398911K [0x0000000680000000, 0x00000006c0000000, 0x00000006c0000000)
  eden space 838912K,  35% used [0x0000000680000000, 0x0000000691f2fd48, 0x00000006b3340000)
  from space 104832K, 100% used [0x00000006b3340000, 0x00000006b99a0000, 0x00000006b99a0000)
  to   space 104832K,   0% used [0x00000006b99a0000, 0x00000006b99a0000, 0x00000006c0000000)
concurrent mark-sweep generation total 4194304K, used 1244832K [0x00000006c0000000, 0x00000007c0000000, 0x00000007c0000000)
concurrent-mark-sweep perm gen total 240952K, used 144550K [0x00000007c0000000, 0x00000007ceb4e000, 0x0000000800000000)

The Java arguments we are using to start the tomat service:

export JAVA_OPTS='-server -Xcomp -Xbatch -Xss1M -Xms5G -Xmx5G -XX:+UseConcMarkSweepGC -XX:+CMSIncrementalMode -XX:NewSize=1G -XX:MaxPermSize=1024M -XX:CMSInitiatingOccupancyFraction=80 -Djava.rmi.server.hostname=192.168.129.122 -verbose:gc -XX:+PrintGCTimeStamps -Xloggc:/opt/Alfresco/GC.log -XX:+PrintGCDetails -XX:+UseCodeCacheFlushing -XX:+CMSClassUnloadingEnabled'

Appreciate your valuable help.

[mrogers]

There should be more information, can't help without it.


For a start do you think its alfresco shutting down unexpectedly or is it your O.S. (which one) shutting alfresco down due to insufficient resources?

[anndrewtr]

At this moment I am not sure whether it is due to Alfresco or OS.

Hereby I have attached the log file for reference.

Kindly let me know how shall I dig further to find the root cause of what is creating the problem.

Any help is appreciated.

[mrogers]

It looks like a 64 bit JVM issue  (or possibly the JVM is incorrectly masking a problem in native code)   If you google for SIGSEGV and 6.0.24 it seems that others have the same problem.   Including a couple of bug reports from Oracle.    There are a few reports that switching JVM back to 1.6.0_07 helps.

[anndrewtr]

Thanks for your reply.

I have looked into the system log and found the below messages:

Jan 3 04:38:54 alfresco2 abrtd: Directory 'ccpp-2014-01-03-04:38:28-45279' creation detected
Jan 3 04:38:54 alfresco2 abrt[48158]: Saved core dump of pid 45279 (/usr/lib/jvm/java-1.6.0-openjdk-1.6.0.0.x86_64/jre/bin/java) to /var/spool/abrt/ccpp-2014-01-03-04:38:28-45279 (6204538880 bytes)
Jan 3 04:38:59 alfresco2 kernel: uhci_hcd 0000:01:00.4: Controller not stopped yet!
Jan 3 04:39:04 alfresco2 abrtd: Sending an email…
Jan 3 04:39:04 alfresco2 abrtd: send-mail: error while loading shared libraries: libmysqlclient.so.16: cannot open shared object file: No such file or directory
Jan 3 04:39:04 alfresco2 abrtd: Error running '/bin/mailx'
Jan 3 04:39:04 alfresco2 abrtd: 'post-create' on '/var/spool/abrt/ccpp-2014-01-03-04:38:28-45279' exited with 1
Jan 3 04:39:04 alfresco2 abrtd: Corrupted or bad directory /var/spool/abrt/ccpp-2014-01-03-04:38:28-45279, deleting
Jan 3 23:21:55 alfresco2 auditd[3321]: Audit daemon rotating log files

Hope this is JVM installation creating problem!!!