I'm unable to sleep tonight so I thought I would push this post out VERY early. When you don't sleep your mind takes interesting turns, which can be a good thing.

I was watching a briefing today by a couple of friends as they were talking about various ways to arrange a Windows Server Cluster for SQL Server. I often see an "active" node of a cluster with a "passive" node backing it up. That means one node is working and accepting transactions, and the other is not doing any work but simply "standing by" waiting for the first to fail over.

The configuration in the demonstration I saw was a bit different. In this example, there were three nodes that were actively working, and a fourth standing by for all three. I've put configurations like this one into place before, but as I was looking at their architecture diagram, it looked familar - it looked like a RAID drive setup! And that's not a bad way to think about your cluster arrangements. The same concerns you might think about for a particular RAID configuration provides a good way to think about protecting your systems in general.

So even if you're not staying awake all night thinking about SQL Server clusters, take this post as an opportunity for "lateral thinking" - a way of combining in your mind the concepts from one piece of knowledge to another. You might find a new way of making your technical environment a little better.