Jeff,

I think your right - the fact the system runs OK during the day is just brute force - a massive overspec of the base system. Generally the system bumbles along on 4 to 6 CPU's active (using top) with the rest idle/sleeping. It may not be nice, but it works.

(This is a basically a government procurement - of course we got screwed eh?)

Will arrange for the redo drives to be reconfigured at the next scheduled down time (need to change control etc etc etc) and I'll see if it has any improvement on the flexibility of the system.

In answer to my original question - Would you expect a CTAS to cripple this system? - I think the answer is a definate no, and I'm glad everyone else agrees. I thought I might be going mad!

Anyway - got to get onto setting up a SQLServer cluster with replication. Oh Joy!