11-29-2010 02:10 PM
When backup job is working it's way down a selection list and comes to a clients that is OFFLINE, the dedup job hangs and will not skip to the next client and continue. This works when going directly to tape or B2D2T.
Solved! Go to Solution.
12-13-2010 01:43 PM
11-29-2010 02:21 PM
This is a PERFECT case for what I've been preaching for a long time.. Create selections lists PER SERVER. This way machines that are offline for whatever reason, the jobs can be put on hold as needed, and machines that are offline do not cause an entire job to fail.
I dont have an answer to your solution, but creating individual jobs/policies per server has been something I've been doing for a long time to really improve my successful backup percentages, as well as ease maintenance tasks.
If doing client-side dedupe you have to have individual selection lists, but media server dedupe does not. Which sounds like you are doing the later.
Now, there was a critical hotfix that came out in early Nov for BE2010 R2 targeting specifically dedupe reliability. Do you have any of that installed? If you do, have you re-pushed out the RAWS as well?
11-29-2010 07:39 PM
Create selections lists PER SERVER.
... but this have to be balanced against the extra administrative chore to schedule the various jobs, updating the holiday schedule, etc.
11-29-2010 10:43 PM
Perhaps. But backup to disk and send multiple jobs to disk concurrently. Who cares about windows? The whole point of dedupe is reducing backup windows as well as storing less data on disk, and leveraging disk for faster restores.
I see this as no different in terms of management. After the initial setup, you only get notified of errors. You would get the same amount of errors to troubleshoot vs individual lists. We're just operationally choosing a different method to reach our end goal. Which is to have successful backups.
Like I said, there is some initial setup, but the long term maintenance out weighs the up front effort.
12-13-2010 01:43 PM