[horde] memcache session handler

[liamr@umich.edu]

We switched to the new memcache session handler this weekend, and it 
was fine for a couple of days, but then exploded spectacularly 
yesterday afternoon.  We had 15 servers defined (a small instance of 
memcache running on each of the webservers in our webmail pool).

The load on the machines were each less than 0.5, but the application 
became super unresponsive in the browser. If users did get an error, it 
was "No connections available for memcached SessionHandler." (which 
looks like it's in _getConForId()).  Sometimes, after spinning for 
several minutes, users would get dropped in the horde portal screen.

Since we had to turn the log level way down to prevent the session 
handler from core dumping, I don't have a lot to go on.

Any ideas what would trigger this situation and how it could be avoided 
in the future?

Liam