Tuesday, June 05, 2012 3:45 PM
We've started to notice LSASS running high (between 40-90%) on our global catalog servers. We've run the Active Directory Diagnostics and are seeing the NSPIBind running high on the processor (57% out of 82% total processer %). This reminded us about the NSPI caching and didn't know if it applicable anymore on Exchange 2010. We have a few options here but wanted to know if anybody is experiencing the same issue. These are the options we are possibly looking at:
MaxConcurrentAPI (we have not really seen any issues):
We are currently running on Exchange 2010 SP2 RU1 with Window 2008 R2 SP1 DC/GC, around 25k users and 3500 OABs.
Monday, June 11, 2012 4:28 PMHow many GC's? What is the ratio between Exchange mailbox cores and GC cores?
Tuesday, June 12, 2012 1:50 PM3:1 mbx to gc core ratio. We added 6 more GCs and it calmed everything down but we are used to seeing sub 10% utilization on our Exchange 2007 GCs so we are unsure if this is to be expected or not.
Monday, June 18, 2012 6:07 PM
We've gone ahead for testing applied the NSPI Caching fix and KB 2545833 on 1 GC which has yielded no change for that particular server. We're still reviewing our performance reports and capturing TCP dumps to see if we can pinpoint anything but we have not been able to as of yet.