One of my clients has raised an issue which is affecting some but not all users. All are on Win 7/ Outlook 2010 - Exchange 2010 SP1 RU3. All are in the same physical site, spread over different databases in different DAGs.
The symptom is slow mail delivery to the full Outlook 2010 32bit and 64bit client. This started on Monday.
I can watch a test email hit the Exchange server mailbox database for a user, and also watch the item count change for the user mailbox through the EMS. The user also receives the message in OWA at the same time and on his BB.
However, the Outlook client does not receive the message; this can range from a few minutes to up to an hour. A manual Send/Receive does not help and the Auto Connection test shows the client is being fully connected. Sometimes the mail will come in a burst of 5 or 6 messages showing a time range of up to an hour.
I figured the next thing to do would be to enable diagnostic logging on the Outlook client and read the contents of the OPMLOG.log. By doing this, i should be able to view transport requests made by the Outlook client and what it sees in the mailbox database for the user. If i know it's already in OWA and on his BlackBerry, then the outlook client should show that the message has been recognized on the server and then being downloaded.
However, when i compare the time received in OWA etc., the Outlook OPMLOG.log shows no messages as having been found.
Can someone point me in the direction of further troubleshooting here, or am i missing something in the way OPMLOG.log displays server requests. If the item is in the database, why wouldn't the outlook client see it?
What's happening on the MAPI connection with the RPC endpoint that i can look at, which should show any latency or communication issues? The CAS servers are load balanced with the same VIP.