As many of you probably noticed, we had two major outages the last two days. I noticed because I got phone calls from our monitors and from Jive at 2:30 am and 6:30 am.
Background on the issue is that it was a memory/disk space issue. It was caused because we requested more frequent database dumps for analytics. The issue should be resolved, as Jive has added an additional node just for the analytics this morning.
I think the nodes fell over in a domino type effect. So if you were on Node 1 you got bounced out or to another node (which would ask you to sign in again). Eventually they were all down and restarted.