Apache httpd dies for no obvious reason.

( Oracle 9iAS on RedHat 6.2 w/ /usr/local/jdk1.2.2/bin/java -version java version "1.2.2"
Classic VM (build Linux_JDK_1.2.2_RC4, native threads, sunwjit))

Once in a while the HTTP listenneer dies giving the below entries in the error log file.

First there are alot of these, which is normal:

[Fri Mar 3 10:02:56 2006] [debug] opm_ew.c(253): OPM: EW enters process
monitoring loop
[Fri Mar 3 10:02:59 2006] [debug] opm_ew.c(253): OPM: EW enters process
monitoring loop

Then something bad happens and the HTTP listener decides to exit:

[Fri Mar 3 10:03:01 2006] [info] OPM: EW is exiting with sig=1073929500

Then it starts killling the child processes (which is too normal):

[Fri Mar 3 10:03:03 2006] [warn] child process 16997 still did not
exit, sending a SIGTERM
[Fri Mar 3 10:03:03 2006] [warn] child process 16998 still did not
exit, sending a SIGTERM

etc..

I've reviewed the access log file for suspected entries. For example if a client somehow called a buggy app I should see the same call in the access log before it dies but I can't find any reasonable pattern.

The HTTPd dies a couple times per 24h, the system resources are not heavily utilised.

Any info on this would very appreciated, thanks.

Rgds

/PL Received on Wed Mar 08 2006 - 12:33:16 CST