I have a 2 node active-passive Exchange 2007 SP3 SCC cluster. both nodes running Windows Server 2003 64 bit. One of the nodes seems to be working just fine. The other is not so healthy. When the other is the active node, users experience a significant performance hits for 15-30 minute periods at least twice a day. I ran performance monitor and collected counters described in the following link while both machines are active and compared the results. http://technet.microsoft.com/en-us/library/bb201689(EXCHG.80).aspx I found the results to actually be similar for both nodes while each were active. some RPC latency values seemed to spike a bit high, but this was consistent on both nodes The one thing I noticed that differed was that store.exe on the "good" node was taking 9gb of memory, while on the "bad" node it was only consuming 4gb, and using Virtual memory for the rest of its operations. Both nodes are configured identically, and no other applications are taking up the missing memory. Store.exe on the "bad" node is simply not taking enough memory. Could this just be a result of bad memory which can be solved by a hardware replacement (if so, what counters can I look at to verify memory is bad), or are there settings I can check on the "bad" node that may be limiting the amount of memory store.exe can allocate to itself?

Hi czaluzec, Could you please run the Exbpa & Performance Troubleshooter in the Toolbox first and post the results here?Please remember to click Mark as Answer on the post that helps you, and to click Unmark as Answer if a marked post does not actually answer your question. This can be beneficial to other community members reading the thread.