Veritas-bu

[Veritas-bu] Reset NetBackup 3.4

2004-05-30 21:38:43
Subject: [Veritas-bu] Reset NetBackup 3.4
From: ccollier AT classmates DOT com (Chris Collier)
Date: Sun, 30 May 2004 18:38:43 -0700
Have you tried to kill the jobs using vmoprcmd? If not, try this:

vmoprcmd -d pr (to find pending job id's)
vmoprcmd -deny <job id>

You might need to add the '-h <media database host>' variable if you have more 
than one server in your environment.  We've been through the same problem and 
if memory serves me correctly this is how we fixed the lingering bpbackupdb job 
issue.

Hope this helps.


On Sun, May 30, 2004 at 09:54:25AM -0700, Chris Pratt wrote:
> Hope someone can help me out here.  NBU DB tape accidentally got swapped out 
> when operator swapped tapes on Thursday.  Thursday night backup worked fine, 
> but failed when it tried to write the db to tape.  I found that I had locked 
> processes and no longer a db tape available.  These locks were preventing 
> daily backups from proceeding until the db was backed up.  I changed the 
> config to use which tape for db backups, stopped all deamons, cleaned the 
> existing locks for bpdbm and bprd, and restarted nbu.
>  
> Somewhere this information is saved because I come right back to the same 
> lock up, the only difference is that the timestamp of the files is now 
> current.  Where does NBU see that these jobs were not completed and how can I 
> *remove* them?  I can't put the original db tape in (not sure that would even 
> help) as I am far off-site.  I need to get rid of all the previous instances 
> of failed backup so nothing is running, backup my db to the tape I assigned 
> from the robot, and requeue my full schedule.  Everytime I clean up the old 
> stuff, it just reappears at NBU start.
>  
> Any help would be appreciated.  Thanks.
>  
> Chris
>  
>  

-- 
Chris Collier
Unix System Engineer
Classmates.com
Desk:   425.917.4730
Cell:   425.442.5072
Email:  ccollier AT classmates DOT com

<Prev in Thread] Current Thread [Next in Thread>