CA, Web serevr have 8GB of ram, quad P4 hyper thread procs.

This site is published via content deployment jobs; 7 GB takes 14
hours to deploy.
I have confirmed that SQL is okay resource wise.

When there is a IIS reset the CPU spikes @ 100% for 15 minutes and the
site will not render during that time.

There some customizations to the master page and web parts.

This site itself is smoking fast the pages render very fast.

When the incremental CD jobs run the IIS working process will consume
2.5 GB of memory and the box starts paging.

IIS is not optimized @ the app pool level but I am certain the problem
is far greater than leveraging the procs for the web garden to 4.

Any suggestions how  best to troublshoot this or any insight would be
greatly appreciated.

Have you already applied SP1 and Infrastrucure Updates to both MOSS
and WSS?
These things fixed lots of misc memory and process management issues,
many of
which are somewhat obscure to figure out. I would start there if you
have not already
applied them.

On Oct 27, 11:55 am, Hulicat <dennis_A_wh...@yahoo.com> wrote:

I dropped the production content database on a virtual and connected
to a new web app and it was fine.

After that I thought I had a someone wrong with my web\app server, I
installed another MOSS server with the current build and added it to
the farm. It is alot faster now.

On that machine I configured the app pool "Web garden" to leverage all
4 procs; I dont know if that was the fix or not.