expected volume 'xxx' got NULL

[asimpson]

I have NetWorker 6.1.2 running on a Solaris 9 OS. We have four 9840 tape drives in a StorageTek Powderhorn 9310 Silo. I currently have a tape drive disabled because a backup was writing to a tape, and then media verification failed:
3/19/03 17:20:31 nsrd: media notice: 9840 tape 000190 on /dev/rmt/1cbn is full 03/19/03 17:20:31 nsrd: media notice: 9840 tape 000190 used 37 GB of 20 GB capacity 03/19/03 17:22:07 nsrd: media warning: /dev/rmt/1cbn opening: I/O error 03/19/03 17:22:07 nsrd: media warning: verification of volume "000190", volid 2027369217 failed, read open error, I/O error 03/19/03 17:22:07 nsrd: media notice: verification of volume "000190", volid 2027369217 failed, volume is being marked as full.

Next it tried to mount a new volume, but it got a read open error, I/O error and said "expected volume 'xxx' got 'NULL'." It continued to attempt to mount new tapes until it encountered too many errors and disabled itself.

I have seen read open error, I/O error on the tape drives before but I have always been able to fix them either by restarting NetWorker, running nsrjb -H, or by reconfiguring the jukebox. This time none of the above helps. I have cleaned the tape drive, same thing happens. I am able to mount a tape through ACSLS just fine, but from the Legato server, I am not able to tar to the tape, it says tar: /dev/rmt/1cbn: I/O error. When the tape is mounted through ACSLS and I do mt -f /dev/rmt/1cbn status on the Legato server, it says no tape loaded or drive offline. I have a case open with Legato, but wanted to see if anybody else had any thoughts as well.

 

[TDun]

Check on/in the9310. Does it have a volume on the drive?
If not, then does Networker think it does if you select the autochanger, edit and look at volumes (loaded fields)?
If it does, then the easiest way is to load the tape into the drive physically, stop and start networker.
The alternative is to remove the entries for the relevant drive from the loaded fields (do not delete the line though). As this needs a lot of care, I would not take this approach unless you are confident that you have made a copy of the file (so can regress) and are happy that you may need to delete the autochanger, recreate and repopulate with tapes.

 

[asimpson]

Thanks for the response. Legato has suggested that this problem is because of a bad fibre card, which we will be replacing this evening. We tend to agree with them since those fibre cards usually only last a few years, and yes I was able to mount and unmount a tape through ACSLS, but not even the OS on our Legato server is able to write to the tape without getting an I/O error which leads us to believe that communications aren't making it from the server to the drive.