Skip to main content

Backup of exchange server took an unusual amount of time

Thread needs solution

We ran into an issue yesterday with people having trouble accessing e-mails in the morning. Direct connected Outlook workstations would freeze intermittently. Cached mode Outlook people would experience significant delays synchronizing their mailbox. Pings to the mail server resulted in 75% packet loss and the few pings that would get through were well over 200ms on the LAN. Trying to do ANYTHING on the Exchange server was next to impossible.

Turns out the snapshot was consolidating at 95% when people were getting into work in the morning. This is unusual and never happened before. The mail server backup job is scheduled to start at 00:00 (midnight). By 5:07 AM it was completed as far as Acronis was concerned, but VMWare took an additional 4 hours to consolidate the snapshot afterwards. Since Acronis was done I opened a ticket with VMWare (12242351911) regarding snapshot consolidation. I sent them the logs from the ESX 4.1 Update 3 host.

I did move the backup time for our mail server back a bit to start at 22:00 hours now.

What I do notice in Acronis vmProtect is a big difference in time between the backup on 11/7 and 11/8.
11/07/2012 00:00:02 - Task 'MAIL' changed its state from 'idle' to 'running'.
11/07/2012 00:33:40 - Pending operation 153 started: "Creating partition image".
11/07/2012 05:06:01 - Pending operation 153 started: "Saving partition structure".
11/07/2012 05:07:13 - Task "MAIL" succeeded.
5 hours, 7 minutes, 11 seconds

After Acronis is done then it tells VMWare to consolidate the snapshot. As per VMWare's email below, the snapshot size was about 410 GB.

Compare that with today's backup...
11/07/2012 22:00:02 - Task 'MAIL' changed its state from 'idle' to 'running'.
11/07/2012 22:17:53 - Pending operation 153 started: "Creating partition image".
11/08/2012 01:24:17 - Pending operation 153 started: "Saving partition structure".
11/08/2012 01:24:53 - Task "MAIL" succeeded.
3 hours, 24 minutes, 51 seconds

I looked at vmware.log on the host at and it looks like it’s the same 410GB as the night before. Still the job was about 2 hours faster for whatever reason. Any idea why the vast difference in job time? Its another weeknight so theres no difference in how many backups are running. The schedules are the same between Monday through Friday.

Here is what VMWare support had to say:

I have looked over your support logs and your system looks healthy, there is no storage issues, looking at the progress of the snapshot consolidation there are no errors along the way it was just slowly getting consolidated.

One thing I noticed is that it looks like the snapshot being consolidated was pretty big, it was around 410GB. Depending on the speed of your storage and if there are no other machines using it the rough estimate is that it will take 1 hour to consolidate 75GB of the snapshot data so it would have taken 6 hours to consolidate this. This is just a rough estimate, if storage is experiencing high latency it will easily take twice this amount.

The only thing I can recommend is to monitor the size of the snapshots and keep in mind that it can take roughly 1 hour to do 75GB of data. If there are other jobs running at the same time like backups it will take longer. Let me know if you have any questions, thank you.

Size of the snapshot
Nov 07 08:51:14.549: vmx| DISKLIB-LINK : Opened '/vmfs/volumes/4a941feb-c523ee7a/MAIL/MAIL_1-000002.vmdk' (0x10): vmfsSparse, 860087970 sectors / 410.1 GB.
Nov 07 08:51:14.572: vmx| DISKLIB-CHAINESX : ChainESXOpenSubChain: numLinks = 1, numSubChains = 1

0 Users found this helpful

I'm curious how you end up with a 410gb snapshot after 5 hours. How much data is on the server and how much activity is there? Are any programs or maintenance tasks running during the backup window?

Hello KJSTech,

Thank you for posting your issue in our forum. I am happy to assist you, solving your problem.

Please ensure that you are using the latest available build of Acronis vmProtect 8. The current build is 8128. This article describes, how to download the latest build.

Such a big snapshot file is unusual and I recommend to manual cleanup the snapshot. This article from VMware knowledge base describes this step:

Committing snapshots when there are no snapshot entries in the Snapshot Manager

If committing is not successful or needs too long, you can also delete all snapshots for this machine: Delete a Snapshot

Please monitor the size of snapshots during next backup with Acronis vmProtect 8. If the snapshot is not deleted after the backup, please contact our support. You will find your available support options in our Customer handbook.

Let me know if you need additional help please.

Thank you.