bpdbjobs -clean hangs NetBackup

[Anton_Panyushki]

Our Netbackup environent is pretty big. We run Solaris-based master server and tens of media servers The vast majority of them are Sun boxes too.
We noticed that once bpdbjobs -clean has been launched, NetBackup stops submitting new scheduled jobs to queue and manual backups don't appear in queue as well. 
On top of that as Netbackup cleans outdated jobs, sar -b shows sharp rise in cache buffers reads. Obviously it is a result of jobs files deletion. Other Solaris perforance tools (vmstat, iostat) display quite normal kstat values.
We suggest that NetBackup locks bpdbjobs functionality while purging old jobs. As a result new jobs don't appear in Activity Monitor. 

We are pondering to kick  bpdbjobs -clean 2 or 3 times  a day via cron to alleviate its negative impact.
  
  

[Mouse]

Have seen this problem many times with 6.0 environments.
From small (5-6 media servers) to medium (40-50 media servers) it has a very slow internal communications between daemons in heavy jobs like that.

If it's happening with 6.5.3 or 6.5.4 I would suggest to open support case for it.
 

[Anton_Panyushki]

Is threre any formal resolution to this issue in 6.0 (tuning etc..)?
I mean is it worth upgrading to 6.5.4 to tackle it?

[Mouse]

Our license renewal service includes free upgrade to the latest patchlevel for NBU for our customers, so, if they are ready, we usually do the upgrade.
As of my experience, 6.5 branch is way better than 6.0 in terms of performance. I've been told by backline engineers that a lot of internal communications stuff have been optimized.