Open Stack

Excerpts from John Dewey's message of 2014-08-14 09:53:50 -0700:
> Interesting.  I haven’t seen that happen before, and have setup memcached in a similar fashion as you.  The memcached client consistently hashes each key across your set of active memcached servers.  So killing memcached on server x, should only result in failures for data stored on server x.  However, the client should account for this and rehash.  
> 

The issue is that there is a timeout that must be reached to detect the
dead server, and occasionally we retry it. If we are recreating the
client each request, that timeout is ALWAYS applied before a server is
marked as dead.