Huge change rate detected by Veeam on Exchange VM

[ddelly]

First time doing this so here goes.....
Im having a weird thing to happen every weekend. We perform and entire VM replication of our Exchange server every evening. Daily change rate is anywhere from 8 - 12 GB. All of a sudden between Saturday and Sunday job, the change rate jumps to over 150GB. Our entire Mailbox Information Store is only 145GB. Do you know what would account for Veeam, determining such a high data change rate. It occurs only between the Saturday and Sunday replication job, meaning the Sunday job report records the drastic change. It started occurring a few weeks ago. Any insight would be greatly appreciated.

[Gostev]

Database maintenance process?

[ddelly]

Thanks G, I'll have a look...
If any further thought are out there, feel free to chime in!

[Frosty]

I concur with Gostev. I get similar symptoms with my Reverse Incremental backups. 6 days out of 7 the incremental backup is less than 10GB, but on a Sunday night it can be up to 60GB or more. When I tracked this down it was due to the Exchange database maintenance routines.

[ddelly]

Thanks Frosty!
I reviewed of the ESE Maintenance schedule. THE schedule runs daily 1am to 5am & the database maintenance (24x7 ESE scanning) option is checked. So I don't know if the maintenance is causing the change... I'll search on for what it might be. I thought it may have been some type of ESX maintenance, but I have 12 VM and this abnormality only occurs with the Exchange server replication.

[foggy]

Defragmentation or antivirus check by any chance?

[ddelly]

No & No unfortunately.

[Vitaliy S.]

Just a wild guess - Is it a Windows server 2012 R2? Have you enabled a dedupe feature by any chance?

[ddelly]

HI its Server 2008, no dedupe features are enabled. I've reviewed logs but cant yet find what might be causing it, Especially b/c it hadn't happened prior to this. I'll continue to dig in & be patient.

[Vitaliy S.]

Hmm...reviewing Windows Event log might help also to determine what task is scheduled to run on these days.

[masonit]

Hi

Having problem with slow backup on a Exchange 2010 vm. VM has 3 disks
1: OS
2: Exchange database (210 GB)
3: Exchange databases (120 + 100 GB)

Every night we have alot cbt read on disk 2. This morning when the backup was still running I looked at the transaction logs and there was only ~5 GB.
2014-11-18 23:29:39 :: Hard disk 2 (440,0 GB) 154,2 GB read

I can't understand why there is so much cbt read. I know that exchange and cbt is not optimal but this can't be right. Is there a way to find out what all the changes in the exchange database is?

We never this this issue on disk 3 that has the same amount of Exchange data
2014-11-19 07:05:43 :: Hard disk 3 (300,0 GB) 29,7 gb read.

We have other Exchange vms that are way bigger but we don't see any cbt reads even close to this.

\Masonit