Wednesday, March 4, 2009

Time heals all wounds

Unless you bleeding like a stuck pig.

We have a cluster that is going down for the count. Last week my company ran into an issue where the cluster would stop functioning for about one minute. Here is the really weird part, the node would not fail over.

What we saw was a classic windows "queue" problem. The Windows Server 2003 on Node1 would just stop for about 1 minute. Strangest thing I have ever seen. I was on the phone for 7 hours with Microsoft Partners Support and they were just as confused.

I wanted to point it out here. I will add to this posting later with details of whats going on.

-Boston Tech Guy

No comments:

Post a Comment