> From: Marvin Addison > Sent: Tuesday, July 14, 2015 6:33 AM > > Correct. What makes this acceptable in many if not most cases is that the lost > state is SSO state where the effect on the user is to log in again. As failure > modes go, that's graceful behavior.
Arguably true, but still not optimal :). Contrary to what seems to be the average deployment, I also encrypt the cluster replication traffic over the wire, so I perhaps have stricter requirements for perfection than generally considered ;). > Peer-to-peer replication incurs a cost and in my experience the failure modes > of > replication are orders of magnitude worse than anything I've seen with > memcached. Perhaps over time Hazelcast will prove itself both reliable and > fault > tolerant, but it's patently new technology at this point and needs some road > time to convince me. I load tested it pretty heavily including random node restarts and it never blipped. We've been running it in production for about a year and a half and I haven't seen a single problem (knock on wood). We've probably done at least 4-5 rolling updates since then where we pulled a node out of the cluster and then stuck it back in, I'm unaware of any user facing issues or unnecessary re-authentications. In any case, I'm pretty happy with it :), and wouldn't really want to trade it out for memcached. Thanks… -- Paul B. Henson | (909) 979-6361 | http://www.cpp.edu/~henson/ Operating Systems and Network Analyst | [email protected] California State Polytechnic University | Pomona CA 91768 -- 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
