Hi, Yes, Memcached processes can crash but it never happens for us (10 Memcached servers, thousands of requests every second). It's very stable and performant. That said, you can add replication using repcache. You can even use Couchbase as a replacement of Memcached (replication / persistence) for almost the same performances. Best regards, Jérôme
2013/8/6 Andrew Morgan <[email protected]> > On Tue, 6 Aug 2013, Paul B. Henson wrote: > > Memcached is more performant than EhCache globally, but what really >>> makes the difference is the serialization through the Kryo library >>> (instead of the Java serialization). >>> >> >> My concern on memcached is that it shards the tickets, and if you lose >> any given memcached node, users will be affected and have to >> reauthenticate. From that perspective, ehcache, with a fully distributed >> cache, seems preferable. >> > > We use memcache+repcache to mirror 2 memcache instances. However, that is > a bit of a hack. When I get a chance to redo our CAS infrastructure, I'll > probably just use a pool of memcache servers. > > As others have said on the mailing list before, is it really that big of a > deal if people have to reauthenticate? Ideally, your memcache servers will > not be restarting very often. Our CAS config has a 2 hour idle timeout and > an 8 hour hard timeout, so our users have to reauthenticate from time to > time anyways. > > Andy > > -- > You are currently subscribed to [email protected] as: > [email protected] > To unsubscribe, change settings or access archives, see > http://www.ja-sig.org/wiki/**display/JSG/cas-user<http://www.ja-sig.org/wiki/display/JSG/cas-user> > -- You are currently subscribed to [email protected] as: [email protected] To unsubscribe, change settings or access archives, see http://www.ja-sig.org/wiki/display/JSG/cas-user
