Hello everyone, apologies for the lengthy post. I have a crisis scenario at work and I think you guys may be the only ones who can help me. I contacted Blackberry already through Verizon, but they weren't able to help much.
First, some background. I administer an enterprise level Active Directory network, running Exchange 2003 Enterprise as our mailserver. I have ten users with Blackberries using the Blackberry Internet Service with Outlook Web Access Exchange integration. After a recent RAM upgrade on the mailserver, my users began having their email invalidated every night, so they need to revalidate every morning. This is very annoying to my users and potentially a big headache for me (one of these users is the CEO!), so I need to get this resolved pretty quick.
Some things to bear in mind:
1. A full image backup using UltraBac 8 runs every night. The backup job starts at 11:30 PM and finishes around 3:00 AM. Nothing here has changed, this job has been set up and working fine with our Blackberry users for over a year, but in the last two weeks, this issue began. I am suspecting that somehow the backup disrupts the Exchange server and causes RIMs polling attempts to fail. Normally the drives are frozen during backup, but when I unfroze the drives, the issue still occurred. I'd think that would solve it, but alas it was not to be.
2. I noticed that the timestamp on the revalidate mails sent by
is always around 1:30 AM. I know that normal polling attempts occur between 5 and 20 minutes on BIS. Is there a hard cap to the amount of polling attempts that is made? Does anyone know if that cap can be increased?
3. Similarly, the backup takes about 3 hours. Is there a way to reduce the frequency of polling attempts to "stretch" out the polling process?
4. From the very little I've been able to find on this issue, some have indicated that Outlook Web Access needs to be reprioritized as a higher priority process, and then it will not go down during a backup operation. I haven't been able to figure out what exactly this means though, can anyone shed any light on that?
5. The timing of this with the new RAM placed in the server seems too good to be coincidence. Yet, I can't possibly see how a RAM upgrade could contribute to this issue, it seems to be too high level for that. Anyone have any thoughts on that?
Thanks everyone for any thoughts/ideas!