You should not be running Exchange backups and Exchange defragmentation
at the same time.

You should run your last incremental or differential backup
(if you are using them) just before running a defrag, run the defrag,
wait for the defrag to complete, and then run a new Exchange full backup.

Here is the Microsoft KB article that gives you some details on why:

        http://support.microsoft.com/kb/328804

Thanks,

Del

----------------------------------------------------



"ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU> wrote on 02/11/2008
02:29:20 PM:

>
> We've been noticing poor performance from our Exchange TDP backups.
>
> 1) The TDP Client runs 4 Streams (One per E-Mail Storage Group)
> utilizing a dedicated Gige backup NIC.
>
> 2) The Backup Server is a IBM p570 with 2 3Gige Ether Channel bundles,
> (Traffic is going over backup NIC) but they are seeing 4-5MBS per stream
> averaging 20MBS combined.
>
> 3) The Exchange Team is running the Exchange online Defrag on the
> message stores during the TDP backups, which I know has to have some
> impact but have not been able to quantify. I would say stop running both
> processes concurrently, but need convincing data for Exchange Admins.
>
> 4)The Exchange Backups occasionally run in the Rc = 425 Errors which
> I've found a few posts on that point to the Exchange Server being too
> busy to process the E-Mail Backups? They also see a few RC = 50's... Of
> course nothing interesting on the TSM Server side log, other than
> session starts and stops.
>
> 5) I think the TCP Settings on the E-mail Server and TDP cfg are
> default, so there's probably TCP Window Size and Buffer Opportunities.
>
> It would be great for some one to share any convincing data as to their
> performance improvements to Exchange Backups.
>
> Thanks again!