Video Screencast Help

NetBackup master host is swapping like crazy even though memory usage is only at 40%

Created: 01 Aug 2014 • Updated: 29 Oct 2014 | 4 comments


just wondering if anybody came across this and if it might be related to NetBackup. We have a Solaris 10 Server with only a NetBackup Master installation running. I was now contacted by our operations that they keep receiving error messages from their monitoring that there is a high amount of page swaps going on. Yet when they check the memory usage it's usually at around 30% never above 50%. We have a total of 32GB RAM installed which should be plenty. When looking at the running backups I also don't really see a performance degradation, though it's hard to really tell if there is no impact. Apparently the 'issue' has been going on for a long time so it does not seem to be a new behaviour.

Could this be some function/feature from NetBackup, anybody has any smart ideas on how to further investigate this? (My Solaris know-how is rather limited I am afraid)

As always any help would be greatly appreciated!



Operating Systems:

Comments 4 CommentsJump to latest comment

Michael G Andersen's picture

Could that the virtual/shared memory is not a big enough chunk of the physical memory, know the Solaris administrators tuned some shm and other memory setting when I was working with Solaris backup servers.

Think there is a tech note with suggested settings for netbackup

Marianne's picture

Ask server owner to download and install 'top' uility.

They can then keep it runnng in a separate window to keep an eye on resourse usage. It will list top 10 processes.

Some TNs:

NetBackup Master and Media Server processes may run slower if they are limited to fewer than 8000 open file descriptors.  This test runs   'ulimit -n' and checks that the result is at least 8000 on NetBackup   servers.  See   for more information.

Performance of NetBackup Master and Media Servers can be affected adversely if the system is configured with low semaphore limits.  This test checks whether the current semaphore limits are high enough.  See for details.

Supporting Storage Foundation and VCS on Unix and Windows as well as NetBackup on Unix and Windows
Handy NBU Links

jim dalton's picture


You'll probably need to do a bit of a sizing exercise. Are you running netbackup under a project with resources? Advisable to do so.

Do a search for it here and it shows you how to do it : its straightforward but does allow you to  get a better grip on resource controls. If left to default I recall you only are able to get at 25% of your installed memory. I got bitten when I had 16G (25%=4G=enough) and donated half thus 8G->2G = not enough, hence lots of resource problems and jobs falling over.

If theres messages in the log and failing jobs then it would make sense.


Christoph Maier's picture

Thanks for all the tipps. I will talk to my Solaris specialists once he is back from vacation and see what we can do. I'll update this post if we find anything