apache crash after shibboleth2.xml change and shibd restart

[Cantor, Scott]

On 1/5/18, 3:27 PM, "users on behalf of Feld, Michael (IMS)" <users-bounces at shibboleth.net on behalf of FeldM at imsweb.com> wrote:

> After making a change to shibboleth2.xml and restarting the shibd service, apache seems to “crash” for a couple of
> minutes and recover on its own, impacting other static web sites unrelated to shib. This is evident by Nagios checks
> that begin returning a timeout error. There is nothing evident in the apache logs whatsoever, and it always recovers
> after a couple of minutes. Has anyone experienced anything like this before?

You probably are running into long shibd startup times, which are covered ad nauseum on the list, with the workarounds. Apache is probably waiting on a response from the daemon to handle a request while it's busy reinitializing.

I doubt whatever you're doing requires restarting shibd so it's a self-inflicted problem.
 
-- Scott