Video Screencast Help

Server backs up way too much data after power loss

Created: 11 Aug 2014 • Updated: 11 Aug 2014 | 1 comment

Our current infrastructure contains three storage arrays and three VM hosts running VMSphere 5.1. In another off-site location to our main infrastructure is where our back up server is. We're currently running Backup Exec 2010R3. We have fulls that run once a week with incrementals on all the other days.

Twice now (in a two year span, so not that bad), we've lost power to all our array/hosts and they went down hard. The next day, we had our incrementals run and they backed up an excessive amount of data. As in more than the normal full amount, and more than exists on the server. Example: A normal full run on one of our servers is 98GB, the incremental after the power loss was 306GB. The server only has 280GB allocated across multiple drives (we don't backup all the drives on this server).

This issue gets fixed by having a full run, but I'm still curious as to what is causing this. Looking at the logs, the job summary information doesn't seem to be doing anything out of the ordinary so I'm not even sure where it's getting the extra data from. 

Operating Systems:

Comments 1 CommentJump to latest comment

VJware's picture

Possibly, something is causing the archive bit/modified time to get affected if backing up physical machines, and if backing virtual machines, the CBT may be getting reset causing the application to lose track of the changes and backs up the data as a full instead.

If you compare the job logs, is there no difference in the amount of data backed up for each resource ?

Would you attach or PM the job logs as well. Thanks.