Unsolved

This post is more than 5 years old

194 Posts

669

February 8th, 2006 05:00

nwrecover error: SYSTEM error: Bad file number

Greetings,

I'm now getting the following error when I try to run nwrecover:

csubds5:/> /usr/opt/networker/bin/nwrecover -s solegato1.xxxxxxxxx.edu
nwrecover: lookup failed to server solegato1.xxxxxxxx.edu
nwrecover: SYSTEM error: Bad file number
nwrecover: Unable to find or browse backups from this client.

csubds5 is the storage node which is Tru64 V5.1B PK4.
solegato1 is the Legato server node which is Red Hat 3.
Legato is V7.2.1

I checked the knowledge database and didn't find anything. I tried this:
csubds5:/> /usr/opt/networker/bin/nwrecover -s solegato1.xxxxxxxxx.edu -c csubds5.xxxxxxxxx.edu

It starts but there's nothing in the database. Sounds like a corrupted database for this client since I'm able to browse other clients. What needs to be done?

Thanks,
Vic

6 Operator

 • 

14.4K Posts

 • 

56.2K Points

February 8th, 2006 06:00

As far as I can see your first error is:
"nwrecover: lookup failed to server solegato1.xxxxxxxx.edu"

That means your client is not able to resolve or connect your server. I would start from that.

2 Intern

 • 

1.1K Posts

February 8th, 2006 06:00

Try using:

nsrck -L6 clientname

- to see if you can repair the index. If not recover the index from backups using:

nsrck -L7 clientname

You could also try checking the media database:

nsrim -X

194 Posts

February 8th, 2006 07:00

dhamson2,

Looks like nsrck -L1 fixed it:

[root@solegato1 root]# nsrck -L1
nsrck: checking index for 'csubds5.xxxxxxxxx.edu'
nsrck: /backup/nsrln/nsr/index/csubds5.xxxxxxxx.edu/.nsr is missing; re-creating

After this nwrecover works for that client.

Thanks
Vic

6 Operator

 • 

14.4K Posts

 • 

56.2K Points

February 8th, 2006 13:00

Hi Victor,

This is really strange. .nsr in index directory is nothing else then skip directive. L1 does check index headers, but then the question is why your .nsr was removed in the first place. But it seems like your index was slightly corrupted indeed and that L1 did the job. The reference to your error can be found in nsrinfo manual. It says there "The index server is running, but was unable to process the query." (another reference is in TB 374, but that one probably doesn't count here as cause was something else).

Usually there is no reason for index to get corrupted under normal circumstances so unless you had crash or not so clean shutdown recently one file system check against /backup file system would be a good thing. You can also grep daemon.log for word missing to see if any other index was affected.

0 events found

No Events found!

Top