we are running 7.5.1 on Win 2k3SP2, our Exchange 2007 passive node is running on Win 2k8 Sp1. The module we are using is NMM 2.2.
We are having failures occurring in our backups which seem to indicate aproblem with Exchange VSS writer, in the event log we get Event ID 2034. this is what MS say about the error:
MSExchangeRepl 2034: The VSS writer of the Microsoft Exchange Replication service found an error when it completed the final stage of the VSS backup operation. In this case, backup will fail. Although the event in the Application log states the specific reason for the error, general causes may include the following conditions:
* The VSS writer received the wrong backup ID.
* The VSS writer encountered an error when the program called in to the VSS framework.
* The VSS writer could not communicate with the active instance of the databases in the backup set.
This error occurs in the final stage of the VSS backup process when the VSS writer checks to see whether the backup application has successfully created and verified the integrity of the backed up files. This Error event is logged when the backup application informs the VSS writer that there are no errors. But then the VSS writer finds an error when it truncates necessary log files and updates the database header with the current backup status.
Unfortunately there is nothing else indicative out there as to how we can overcome this problem as in Win 2k8 VSS can't be switched off and ofcourse we need VSS with NMM anyway.
Do the Networker and Exchange server have a firewall in between them?
I used to have problems with jobs running to completion only to have it error out on the end after what appeared to have gone correctly. Come to find out, it was a timeout on the firewall, closed the connection after the standard default of 60 minutes.
For the most part yes we get 100% every night. Mind you, we do our fulls over the weekends. And incremental during weekdays.
Somenight network hiccups may cause a failur of one or two servrs, but others will finish fine. (I should let you know I backup over 300 exchange servers across 3 datacenters.Via firewalls, and all that wonderful networking stuff.)
I should mention we don't use VSS to do our backups. We do the database streaming method. So our save sets are:
MSEXCH:IS/Storage Group 1 MSEXCH:IS/Storage Group 2
And my backup command line is:
nsrxchsv -c
etc.
We are running clusters with 6 active 2 passive. The streaming side works well for us though.
Aside from that, with specifying storage node, and recovery storage node. That is how we do our clusters.
HacknDos1
1 Rookie
•
66 Posts
208
0
Posted September 30th, 2009 12:00
I used to have problems with jobs running to completion only to have it error out on the end after what appeared to have gone correctly. Come to find out, it was a timeout on the firewall, closed the connection after the standard default of 60 minutes.
Thanks
Paul