Networker volume issue

[Martin999]

OK, bit stuck on this ...

I think it's a corrupt media database, and the fix is to delete the problem volumes and relabel. Would appreciate any help though in case I'm wrong.

In nutshell, one particular jb on a networker server refuses to mount a number of its volumes, has a go, then gives error "Expecting volume AB1234, got null".

On perhaps 2nd or third attempt, it might mount the same volume, or might not!!!

Loaded a problem volume, without mounting - and then ran scanner which confirmed the correct label. The volumes are going into the correct drives. All server settings are correct, and the rap.log shows no config changes recently.

Thanks,

Martin

 

[Novelli]

Hi,

question..do you use EMC DL7xx Series as Tape Libary ?
Because we have there similar problems...

If you think its a database problem, try using nsrim -X

And do a inventory of the libary, but before create a output with nsrjb -jlibname and pipe it to a file before the inventory and after the inventory, then do a diff afterwards and look what is different.



Cheers Novelli
Novell Master CNE + CDE

http://www.eckertweb.de

 

[Martin999]

No, the library is an IBM3584 ...

Tried nsrim -x and nsrjb -HE and an inventory.

M

 

[Novelli]

Hmm,

be carfull, nsrim -x and nsrim -X are two totally differnt things !!!

Intresting with your IBM3584, because we simulation the same in our cdl and having similar problems...mabye the virtualization is as good as they simulate errors too ?

Which networker do you have ?

If 7.2.2. try patches

LGTpa 87725
LGTpa 87753
LGTpa 86328
LGTpa 90067 (increases the SCSI timeout to 60 seconds)





Cheers Novelli
Novell Master CNE + CDE

http://www.eckertweb.de

 

[Martin999]

Good point, I meant nsrim -X ...

I don't believe the issue is with the library at all, as the riight volume is going in the right drive, 99.999% sure it's a networker media database issue.

Would have to check the version, off the top of my head I think it's 7.2.1 in this instance. Patching is a fair point, wouldhave to check, though I suspect we are fully patched. That aside, it was working ok ....

M