A couple weeks ago we were doing our server upgrades and went to GroupWise 8.0.3 hp1.
Servers are all SLES 10 SP4 with EXT3 file system.
I've got a total of 4 POs - all are different servers and are only running eDirectory and POA.

Prior to the upgrade the nightly maintenance (Analyze/Fix, message/user, structure) took 64, 32, 14, and 1 minutes each.
After the upgrade the nightly maintenance is now at 230, 86, 14 and 1 minutes each. (The 3x increase on the first 2 is the core of my question)

This has also translated into our backups (are taking twice as long on the first 2 servers, no real noticeable change on the other 2).
The weekend process on 1 server (Analyze/Fix, Message/User, Contents , Statistics, and Update diskspace) took 47 hours to complete (it used to be in the 12 hour neighborhood) - the processor actually spent 122 hours processing the job - so it was using multiple threads.

While the maintenance tasks are running the experience on the GroupWise client is slowed but still usable.
Obviously the amount of data and users decreases as I move through the servers but there's a clear cut line in when the issue started appearing.

Does anyone have any thoughts or run into this themselves?