backup times increased
not really a huge problem (I think) but more of something that might turn into a problem.
The backups have recently started taking a lot longer to complete.
26/07/11 - 8hours 39Mins - 440Gb
02/09/11 - 20hours 49mins - 424GB
I have recently defragged the servers but I would have thought this would have made it quicker if anything, not slower??
We use backup Exec on Server 2k8.
Anyone got any thoughts on this??
Are you backing up to tape or disk? Also are you creating new backups or amending existing archives?
Ah, good point. I should have mentioned...
were backing up to tapes and it is a new backup nightly.
Has anything changed - e.g. location of the data? Is all the kit plugged in where it was before?
Sorry more questions ...
Is the data co-habiting with any other that might be thrashing the source disks?
Any new AV? - scan on access setting can impact backups in some cases
Is the bulk of the data the same?
Say, you haven't gone from nice chunky big files to lots of small files?
Have you checked the health of your arrays?
Expand the job log and see how long each part of the job is taking compared to beforehand... you should be then able to isolate what server is slowing it down (unless it's a general sluggishness across all the jobs)
Another thought is contention at the tape writing stage.
You may need to increase threads/parrallelisms or partition the library to balance demand across devices.
Is it on all tapes or just one or two? How old are the tapes? As tapes get old they can sometimes get slower I've found.