You are on a networker version before 7.3 then. A couple of things you can do;

1. Try to find out the nsrmmd process that is binding the drive. You can do this by issuing "fuser /dev/rmt/1cbn". Kill this process and retry the deletion.
2. You could try to stop the Networker services, clear /nsr/tmp, restart the Networker services and retry the deletion.

Hi,

Have you seen if there is any media mounted on this device?

Have you tried remove this device with #jbedit?

#jbedit ¿d ¿o ¿j nom_jukebox ¿f rd=sunrack2:/dev/rmt/1cbn ¿s backup_server

Look for if this device is being took for Storage Node, and then kill its nsrmmd.

Regards.

Inma Bernabé.

Well, it's simple. nsrmmd is busy with your device. Usually you delete device after library and you do not start that before devices have been unloaded and state of it is ejected. To correct that afterwards you have several approaches (some listed already), but I prefer nsr_shutdown and nsradmin while NW is down

I agree that a jukebox should have all drives unloaded before being deleted.

I recommend trying to unload the drive with "nsrmm -u -f /dev/rmt/1cbn" before killing the nsrmmd.

If everything fails to get rid of the drive busy status, put the drive in service mode, shutdown NetWorker and reboot the server.

I have had to do this if the drive is really wedged and/or the nsrmmd cannot even be killed with a kill -9.

Well, if you can't kill it with -9 that indicated IO issue with device which is different pair of shoes I believe.

If there is indeed a tape stuck in the device, have found that physically ejecting all the drives (open the tape library and eject the drives) and then restarting NW and running nsrjb -HEv worked for me.

Have come across the problem of nsrmmd not getting killed. That is basically because of IO related issues as Hrvoje mentioned. The only way out I have found is to reboot the server.

Yes, sorry I didn't state that earlier. I agree, if kill -9 doesn't work, it's time to re-boot the OS.