On Fri, 8 Mar 2002, [iso-8859-2] Jarecsni J�nos wrote:
> Date: Fri, 8 Mar 2002 17:55:01 +0100
> From: "[iso-8859-2] Jarecsni J�nos" <[EMAIL PROTECTED]>
> Reply-To: Tomcat Users List <[EMAIL PROTECTED]>
> To: Tomcat-User <[EMAIL PROTECTED]>
> Subject: Tomcat 402 Hanging after a few days
>
> Hi,
>
> this week there were some postings regarding $subj. I filed a bug report
> this afternoon, after our tomcat hung again (it was last restarted on
> Monday). One of the Catalina processes dies so completely, that it cannot be
> killed with kill -9. Well so far I thought this can only happen on a
> windows... :)
>
> so check out this bug and vote for it:
> http://nagoya.apache.org/bugzilla/show_bug.cgi?id=6990
>
Voting for it won't do any good, if that is all that happens.
What will do some good is investigating far enough to create a
reproducible test case so that the cause can be understood, and therefore
dealt with. Just to give you a feel for the range of possibilities, the
real problem could be in:
- The OS you are using
- The JVM you are using
- Native libraries you are using
- Tomcat
- Your webapp
Without some help in narrowing it down, there's not much any of the Tomcat
developers can do except mark such a bug as WORKSFORME (and it does - I've
seen and heard about Tomcat running many webapps for much longer periods
of time).
Note: If "kill -9" doesn't work, then there's guaranteed to be an OS
problem -- it should not be possible to create an unkillable process. But
that is undoubtedly not the whole story.
Note: On a Linux box, Tomcat looks like a lot of processes, but it's not
-- those are really threads. So doing a kill on one of the threads is not
going to bring the whole JVM down. Do a "ps --forest" command to see
which is the parent PID.
> ciao,
> J�nos
>
Craig
--
To unsubscribe: <mailto:[EMAIL PROTECTED]>
For additional commands: <mailto:[EMAIL PROTECTED]>
Troubles with the list: <mailto:[EMAIL PROTECTED]>
