Redundant device host showing MISSING_PATH

[Toddman214]

Netbackup 7504, Windows 2008r2 Master and Media servers, with some Linux media servers in the environment.

Hello all.

We had a Linux server fail, which was also running the Netbackup media agent, as it is backing up its own databases through fiber to tape. Once a new server was stood up by Unix support to replace the failed one, and fiber-zoned to the tape drives, I rescanned the server on my end, and I can see the new server (of the same name) and it appears to be running its backups fine. However, I am left with an old path showing in the Activity Monitor under the 'drives' tab as MISSING_PATH:1:0:3:0  <serial number varies depending on the drive path I click>. This does not appear to be affecting the backups, but is causing several of my control paths to show as "DOWN"....no little red arrows or anything, all green.

I do not have access to the Linux servers in our environment, aside from the tweaks available through the Netbackup GUI. Is there some trick to getting this redundant drive showing the "missing_path" to go away from my end?

Thank you,

Todd

[RamNagalla · Accepted solution]

in simple way.. run the Device configuration wizard on the media server once again... it will clear the missing paths 

[Toddman214]

The media server is a Linux server, to which I do not have access. I've reached out to one of our Linux admins to run this. I'll follow up with the results.

[RamNagalla]

thats fine... since you have the GUI access..you can also run it from the GUI--> configure storage devices.. 

[Marianne]

I would personally first delete all devices for this media server from the GUI, and then run device config wizard again - select the robot control host and this media server.

[spitman]

I agree with Marianne. I know that some folks worry about removing robots/drives once they are configured (if it ain't broke, don't fix it) but removing drives/robots and using the device configuration wizard are nothing to be afraid of and, if you have time, is something you should truly be familiar with in case there is some kind of disaster with your library hardware and it needs to be replaced.

Marianne, I think I remember that when you fully remove a robot and drives, you need to do so from more than one place in the GUI--like in the Storage -> Storage Units section, and in the Devices -> Drives|Robots, correct? Is there a third place to check to be sure it's gone?

[Toddman214]

I've used the device configuration wizard many times for recovering after various issues, but generally during a planned maintenance window. Will removing drives kill existing tape write jobs, or just prevent new jobs until the rescanning process is finished? Do I just go in under devices > drives > and delete all of them, confirm, and then rescan? The redundant "missing_path" to this particular media server is seeing all of the various serial numbers for the tape drives, so I'm assuming all drives that can see that redundant box need to be removed and readded. Please correct any misunderstandings I may have. I certainly dont mind. Backups going down in our environment is a very "front-and-center" issue. :\  Thanks for all of the input!

[Marianne]

Apologies for missing your last post.

You can only delete devices and run a scan using the Device Config wizard when no backups, restores or duplications are running.

Deleting ans recanning will both need to restart NBU Media Manager peocesses/daemons which cannot be done while devices are in use.

You only need to remove and re-add devices (tape drives), not Storage Units.