Error when deleting old drives fromGUI

[tigerlover]

We want to delete old drives with missing path using GUI from master server. But got "could not connect to vmd on host" error. I checked both Master and media server the vmd process IS up and running. The physical drive already gone because couldn't see it from library.  How can I delete it? Another question is: Is there any way to push from Master server to all media server and let them  see the newly replaced drive?? If it needs to stop any processes after pushing it will effect any jobs the backup is running now? I am running Solaris 10 on both Master and Media server, NBU is 6.5.6.

[J_H_Is_gone]

Where are you running the console from - I ask because when I run the console from my pc I cannot delete either - I have to run the console from the master server.  I use the windows remote admin - you don't say if you are running that or the java.

look up the tpautoconf -replace_drive command.(not sure if that is avail on 6.5.6 or not I know it is in 7.0.1 so it should be.

[Marianne]

Use tpconfig on media server to delete drive.

tpconfig is a text-based menu.

Remember to restart NBU on media server after deleting the drive.

[tigerlover]

I am running the GUI from my PC which was export the DISPLAY from my Master server..jnbSA&. Was this way working for you? When you use tpautoconf, it will stop some of your running process and restart them? If yes, it will effect my running job? Pause is fine but I don't want the backup jobs which are running now need to be restarted from the beginning because I am backup a pretty big DB. Thanks!

[tigerlover]

I have about 30 media servers, do I need to go to each one to do it?

[Marianne]

"I have about 30 media servers"

A bit different from the reference to master and media server in opening post!

Have a look at this TN:

http://www.symantec.com/docs/TECH34296

DOCUMENTATION: How to update NetBackup for a replaced tape drive without deleting and re-adding the drive

[tigerlover]

This tech note only talked about replacing drives, not about delete drives, when I tried this command, I got the error below even though the path IS there. I ran this from the media server.

/usr/openv/volmgr/bin>./tpautoconf -replace_drive IBM.ULTRIUM-TD2.012 -path /dev/rmt/1cbn
Operation failed, could not find device with path /dev/rmt/1cbn

ls -l 1cbn
lrwxrwxrwx   1 root     root          75 Oct 13 18:10 1cbn -> ../../devices/pci@3,700000/SUNW,emlxs@0,1/fp@0,0/st@w500507630061da7a,0:cbn
 

[J_H_Is_gone]

running x from the master is the way that works.  then I am not sure why you get the error.

[Marianne]

Correct - the TN is about replacing drive without deleting first. Deleting a tape drive requires device management daemon (ltid) to be restarted to take effect. This cannot be done while backups are running.

An entry in /dev/rmt does not mean that the device is really there. The OS does not automatically cleanup device paths.

Pease share output of:

mt -f /dev/rmt/1cbn status

/usr/openv/volmgr/bin/tpautoconf -report_disc

[Omar_Villa]

You can run tpconfig commands under vmoprcmd -h media -devconfig "-multiple_delete -drive drindex,..." do it for each media from the master