[This is something I’ll revisit from time to time and update.]
There are some commands you should always be aware of with NetWorker, regardless of whether you like the command line or not. Here’s some of them:
# nsrim -X
This does a check of the save sets against volumes – while the EMC documentation suggests you should only need to run this after a crash, here’s an alternate policy: you need to run this at least fortnightly, or preferably weekly. (If you are running low on media, or low on space on disk backup units, running nsrim -X can also in a pinch be used to force any recycling that may be ready but ordinarily wouldn’t be processed until around midnight.)
# nsrck -m
This performs another style of consistency check on the media database, and corrects errors when they occur. It actually does cumulative repairs, so when you need to run it, you should be prepared to run it up to 3 or 5 times before logging a support call. When do you run it? After a NetWorker crash – if your server does crash, you really, really need to run this command. It should give you no output – if it gives you output, it means there’s been issues, and you want to run again and see whether all issues have been corrected.
# nsrck -L3
This is a good mid-level check to run against the indices that can clear up redundant entries and act as a good sanity check against indices. It’s also a good starting point if you’re having intermittent backup issues that aren’t network/hardware related, and like nsrim -X, it’s a command I recommend running at least fortnightly, or even weekly.
(All of these commands should be run when there’s no backup activity.)
On a very busy NetWorker server, sometimes nsrim -X doesn’t run. It can be a good idea to run this every few days if you’re noticing tapes or disk savesets hanging around when they should be recyclable…
I recently had a PSE emphatically tell me that nsrim -X is never run automatically by NetWorker, but instead just a regular nsrim is run. I’m still not convinced I believe that, and for one key reason. In earlier releases of 7.x, I had some servers running Linux that would core dump and fail to complete an nsrim -X every time the command was run (nsrim by itself seemed fine though). Every morning a little past midnight though, a new nsrim coredump would be generated.
So I sort of take what the PSE told me with a grain of salt, but agree with what you’re saying – if you’re scratching around for media or space, running nsrim -X can save the day.