Abnormal server crashes: no information provided concerning the crash

Question

Hi,
I got a WLS5.1 SP9 server running on solaris 2.6 with JDK 1.2.2
A "classical" web application is deployed on this server
The problem is that sometimes (and not in a regular way) the server crashes without
providing any exception nor error message: it just crashes as if it was killed,
and I'm sure nobody killed it. I currently trace the memory using the -verbose:gc
option to ensure that it's not a memory problem, and every time the server crashes,
there's still enough space in the heap.
Any idea of what could be the problem, or how to know what's could be the reason?
Thanks in advance,
Cheers,
Stephane

Answer 1

We're experimenting the same problems with different versions of
Weblogic
Weblogic 5.10 SP12 - SDK 1.3.1 - Solaris 8
Weblogic 5.10 SP10 - SDK 1.3.1 - Solaris 8
Weblogic 5.10 SP10 - SDK 1.3.1 - Solaris 2.6
We also tried to add logs with
-verbose:gc
-verbose:jni
-verbose:class
to see if any JNI or specific class call could bring the problem but
nothing interesting was found int the logs.
We used the "truss" command on the Java process but unlucky, the process
never crash when truss is launched (I think it slowers too much the JVM).
"Stephane" <[email protected]> a écrit dans le message news:
[email protected]..
>
Hi,
I got a WLS5.1 SP9 server running on solaris 2.6 with JDK 1.2.2
A "classical" web application is deployed on this server
The problem is that sometimes (and not in a regular way) the servercrashes without
providing any exception nor error message: it just crashes as if it waskilled,
and I'm sure nobody killed it. I currently trace the memory usingthe -verbose:gc
option to ensure that it's not a memory problem, and every time the servercrashes,
there's still enough space in the heap.
Any idea of what could be the problem, or how to know what's could be thereason?
>
Thanks in advance,
Cheers,
Stephane