memory leak?

[Steve Wills]

On Sep 13, 2011, at 9:43 AM, Cantor, Scott wrote:

> On 9/13/11 9:14 AM, "Steve Wills" <swills at webassign.net> wrote:
>> 
>> I had planned a restart, but things went haywire before I could do that.
>> Memory usage shot up from the 600+mb it had been using to about 2.7GB in
>> a short amount of time. Restarting shibd and apache resolved the issue.
>> It's running fine now using about 366mb. Do you have any ideas what could
>> have caused that or how I might go about searching for a cause?
> 
> Yes, you are using prefork Apache and your stacks are too big. Stop using
> prefork, problem solved.
> 
>> I've already looked at the logs and there didn't seem to be a
>> significant increase in number of requests.
> 
> Every Apache request generates traffic to shibd. Every Apache child
> process by definition implies at least one shibd thread.

Thanks Scott, that explains it. Would you recommend MPM event or worker? I'm guessing worker?

Steve

--
Steve Wills
WebAssign, http://www.webassign.net/
Phone: (919) 829-8181 x216 Cell: (919) 622 6826