CiviCRM Community Forums (archive)

Greetings,
We're currently running civicrm and have been deployed now for a few months, and things in general are working great (we need to upgrade grant you);
One issue we've been having with our deployment is that when I go in and do an "Intensive" operation on with civicrm (like generate pdf or some longer running tasks), the whole civicrm deployment will drop off the face of the earth and white screen for about 2-5 minutes.  After that, sometimes the operation will be completed, sometimes it wont, and and everything appears to be running fine again.

The remainder of the wordpress site is running without issue, and the kicker is that the nginx error logs show absolutely no issues at all.

We've been bating around ideas as to what the issue could be, and the only thing which we can think of is somehow the memcached server is crashing.  The server we've moved to in the last few months is more than powerful enough to run whatever we want (32GB ram, dedicated server), so we keep coming back that perhaps its a server configuration issue, possibly with memcached (Wordpress doesn't use memcached, civicrm does in our setup)

Is there any settings for memcached which to use for recommendations, or other potential thoughts as to why a failure like this could occur?

Unfortunatly the error has been too intermittent to be able to enable debugging and see if that even shows up.