/nsr filesystem out of space. need to clean up indexes...

[jesseroscoe]

Hi all,
I'm running Networker 7.1 (I know it's a little out of date) on Solaris 9 and am having space issues on /nsr. When I saw that this filesystem was running low on space, I cut my browse and retention policies down from 1 year to 6 months on my clients. Does changing the browse policy for a client affect how long indexes from old save sets remain, or does it only affect the indexes created after the change is made? Essentially, I am trying to find the best and/or fastest way to remove the index files for each client that are more than 6 months old. Due to the number of save sets and clients, I would prefer to use something other than the nwadmin "remove oldest save set" option, as this would be far too time consuming.

Your help is greatly appreciated. Thanks.

 

[605]

Sorry to say that but this message is obsolete - it refers to a NW 5.x environment.

To compress a CFI DB physically, it was necessary to copy its valid entries. Consequently, worst case, you needed at least the same empty space that was used by the largest CFI. If the empty space on that partition was too small to support that, this notification was triggered.

Since NW 6.0 there was no need for a copy any longer but the message still appeared. Consequently, it can be that you have GBs of empty disk space but still receive this message. If you all consider this, you may carefully ignore this message.


To answer your question:
- Changing the browse/retention policies only affect future backups.
- If you want to shrink the CFI for existing backups, you can also ...
- change the browse time (nsrmm -e browse_time -S ssid)
- delete a CFI for a save set (nsrmm -dP -S ssid)
- set the save sets to recyclable (nsrmm -o recyclable -S ssid)

If you script that, it should ease your task ... if you still want to delete the CFIs ;-)

Of course, you can also move CFIs to another partition. This is probably the fastest way.

 

[jesseroscoe]

Thanks for the info. I'll probably end up having to script it to remove the old stuff. I've already moved some of the clients to a different partition previously when I decided to lower the browse time, but I don't have enough space to move any more. I was hoping that reducing the browse time would have fixed my issue but didn't realize it would only affect future backups. Legato documentation seemed a little vague on that issue.

Thanks again,
Kevin