general/1072: http servers die until only one left in process list

[Robin Laing <ad...@tst-medhat.com>]

>Number:         1072
>Category:       general
>Synopsis:       http servers die until only one left in process list
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    apache (Apache HTTP Project)
>State:          open
>Class:          sw-bug
>Submitter-Id:   apache
>Arrival-Date:   Thu Aug 28 14:40:02 1997
>Originator:     admin@tst-medhat.com
>Organization:
apache
>Release:        1.2.3
>Environment:
Linux 2.0.30 (Red Hat 4.2)
128 Meg Ram
Pentium 90
>Description:
We have upgraded to 1.2.2 and then 1.2.3 this last month and we have had the same problem.

The number of servers dies until only the original server started is running.  A SIGHUP will correct the problem for a short period of time and then it will stop again.

We have set the MinSpareServers to 10 and StartServers to 15.  Once the servers start dieing, the only way to get the dynamic servers to work again is to restart the server using -SIGHUP or kill and a full restart.

There are no messages in the error_log in regards to this problem. 

I could not find anything in any of your bug reports on this problem either.
>How-To-Repeat:
I don't know.
>Fix:
We are now running a cron job that checks the number of servers and if ti drops below a set level, we do an automatic -SIGHUP and this seems to fix the problem for a couple of hours.  Last night (97-08-28) this had to be done 5 times.

%0
>Audit-Trail:
>Unformatted: