cancel
Showing results for 
Search instead for 
Did you mean: 

BE2012 Dedupe Disk Full

FJ1200
Level 3

Hi - Our 17TB dedupe disk is full and has a lot of very old expired backup sets of servers that were decommissioned 2 years ago.  I have spent 2 days reading through every available KB article and forum post I can find relating to this, and have tried all recommendations without success.  And yes, I have read TECH130103. 

I have 2 major issues:

1.  I need to clear the old backup sets out - I have added in the registry DWORD to allow removal of expired backup sets - no difference 24 hours later.  Manually deleted some old sets - still there this morning.  So how does that work?

2. Frequent running of "crcontrol --processqueue" finishes almost instantly, and "crcontrol --processqueueinfo" shows Busy=No, Pending=Yes, and I can't clear the Pending ones.  I think they are tape jobs I have some with the status of Ready: No device available.  How do I clear them?  Putting the queue on hold makes no difference.

Is there a reasn why I can't run the dedupe compression with pending jobs on hold?  I have removed a lot of stale jobs from BE including those that relate to some of the old backup sets.  We now have around 220 jobs.

I have recently taken over responsibility for our backups and this is my first (and hopefully last) experience with Backup Exec.  We may look to upgrade to 2015, but we are migrating to a different environment in the next 3 - 12 months so are not sure if the cost benefit is there and so I have to make this work but it's driving me nuts!

 

1 ACCEPTED SOLUTION

Accepted Solutions

FJ1200
Level 3

We are now in the process of proposing a 2015 upgrade so I'm hopeful we'll get a result.  In the meantime I have been running the dedupe cleardown manually, and do seem to be making progress - albeit painfully slowly.  The processes have taken several days to run and the garbage collection - running now - is going to take time.  Then will run the compression which should give me back 650GB at least according to dsstat.

Many thanks for the advice given.

 

View solution in original post

8 REPLIES 8

CraigV
Moderator
Moderator
Partner    VIP    Accredited

Best to log a call with Veritas themselves to help you. If you have followed the official TN around this and it still hasn't worked (and you can't throw more disk space at the issue), then let support assist you. Best they do it in case there is some sort of corruption, with them fixing it.

Thanks!

FJ1200
Level 3

ok, thanks for the reply.  Will see what they say.

 

 

 

FJ1200
Level 3

Well - having deleted almost 500GB of old backup sets yesterday and left it to run overnight I've come in this morning  to - Guess what?  No Change!!  Still shows as full, still failing backup jobs, still unacceptable. 

There HAS to be some way of getting rid of these expired jobs and reclaiming the space.  I must be missing something somewhere. 

I've double-checked everything, it is possible there's something buried somewhere I'm missing, but I badly need this working.  Our SQL backups are taken care of elsewhere thankfully. 

The main thing I need to know really is can I still run a dedupe compression with some jobs pending if I put the queue on hold or do I have to ensure there is nothing pending?

Can old or expired backup sets be removed with Powershell?  I already use it a lot with BE.

CraigV
Moderator
Moderator
Partner    VIP    Accredited

as I mentioned above, rather log the call with Veritas to get them to assist you. Don't muck around with the dedupe folder.

Thanks!

FJ1200
Level 3

Well - still waiting for a replay from them.

CraigV
Moderator
Moderator
Partner    VIP    Accredited

I'd recommend escalating it, or sending a PM to VJware and see if he can get this done internally.

Thanks!

pkh
Moderator
Moderator
   VIP    Certified
Once this is over you should plan on upgrading to BE 15 which has a lot of fixes for DLM and may solve your problems

FJ1200
Level 3

We are now in the process of proposing a 2015 upgrade so I'm hopeful we'll get a result.  In the meantime I have been running the dedupe cleardown manually, and do seem to be making progress - albeit painfully slowly.  The processes have taken several days to run and the garbage collection - running now - is going to take time.  Then will run the compression which should give me back 650GB at least according to dsstat.

Many thanks for the advice given.