[cobalt-users] re: proper status from RAQ3 admserv and httpd?

[Paul Ferwerda <paul@xxxxxxxxxxxx>]

Running Apache 1.3.20 built from source on a RaQ3. The system had been running fine until sometime in the last month or so. Now, as I mentioned in an earlier message, the webserver periodically stop responding and I have to SSH and shutdown both the webserver and the admserv and then restart them.

I checked the error log and see that it is filled with the following sorts of messages:

[Sat Apr 20 03:05:36 2002] [notice] child pid 6875 exit signal Segmentation fault (11)
[Sat Apr 20 03:05:39 2002] [notice] child pid 6843 exit signal Segmentation fault (11)
[Sat Apr 20 03:05:45 2002] [notice] child pid 6831 exit signal Segmentation fault (11)
[Sat Apr 20 03:12:46 2002] [notice] child pid 6824 exit signal Segmentation fault (11)


earlier in the log there is:
[Fri Apr 19 18:02:20 2002] [notice] child pid 3280 exit signal Segmentation fault (11)
[Fri Apr 19 18:02:23 2002] [notice] child pid 3253 exit signal Segmentation fault (11)
[Fri Apr 19 18:06:59 2002] [error] server reached MaxClients setting, consider raising the MaxClients setting
[[Fri Apr 19 18:07:19 2002] [notice] child pid 3336 exit signal Segmentation fault (11)
FATAL:  erealloc():  Unable to allocate 245760 bytes
FATAL:  erealloc():  Unable to allocate 61440 bytes
FATAL:  erealloc():  Unable to allocate 61440 bytes
FATAL:  erealloc():  Unable to allocate 15360 bytes
FATAL:  erealloc():  Unable to allocate 61440 bytes
FATAL:  erealloc():  Unable to allocate 15360 bytes
FATAL:  erealloc():  Unable to allocate 61440 bytes
/usr/sbin/sendmail: error in loading shared libraries: libnsl.so.1: cannot map zero-fill pages: Cannot allocate memory
FATAL:  erealloc():  Unable to allocate 245760 bytes
[Fri Apr 19 20:10:38 2002] [notice] child pid 3319 exit signal Segmentation fault (11)
[Fri Apr 19 20:10:42 2002] [notice] child pid 3325 exit signal Segmentation fault (11)

I've got at least two :) questions:

1) is there any place other than the error log to find out more details on why we're getting a segmentation fault, i.e. any place that might have a stack trace for example?

2) It looks like the system is running out of memory and not very gracefully.  Any suggestions on what to do short of setting up a cron job to cause it to reboot periodically?  I've got all the cobalt RaQ3 patches applied

Thanks,
Paul