It could be (but hopefully isn't!) that the Java virtual machine is
crashing. If that's the case, then in the domains/domain1/config
directory there will be a file called hs_err_pid which will have
information on what caused the JVM to crash. [At least with standard
JVMs -- I'm assuming CentOS just runs Sun's standard Linunx JVM?]
The JVM should never crash, so if that's what's happened, you should
file a bug against the JVM. Most often, crashes are caused by memory
issues -- have you changed the JVM configuration options in your domain?
If so, you may need to reduce the heap size. But it will really depend
on what the error file tells us.
-Scott
On Mon, 2007-04-23 at 10:18, Lucas Jordan wrote:
> I know this is a long shot, but I am having trouble with an instance
> of glassfish. The app server simply stops once in a while, that is the
> say, the app server goes down and there is no java process running
> when I check on it. This has happened several times, and each time the
> log files in domains/domain/logs/* tell me nothing about what
> happened, they simply contain normal log entries.
>
> I would like some advice on how to debug this. I am running on CentOS
> 4.3 which other log files might help? next time I restart the server,
> can I do something to better track the process?
>
> Thanks a lot!
> -Lucas