Tape Drives duplicates after Tape change

[Stephan23]

Hello everyone,

I noticed a strange behavior since an upgrade to Veeam 8.

There is one backup server with a local repository and an integrated tape drive, which can hold only one tape. The tape job runs every night and has a single Media Pool as a target with all available tapes. Every Morning the tape gets changed to the next one in the rotation.

The problem is, that after the tape change, the drive gets duplicated in the "Tape Infrastructure" view, both of them with the same Drive ID, both show tape media as online (the tape that got changed and the new one). The next day it duplicates again and shows 3 tape drive with 3 different tapes, once I saw up to 5 drives. Sometimes those "fake" drives vanishes without any input in no specific order. Sometimes one "fake" drive shows me tapeX, but when I wait a minute or so it changes to tapeY, if I click somewhere else in the navigation and then back to the online tapes it's tapeX again.

I created a support case about that and was told, that the tape needs to be ejected manually in the Veeam console before changing it. The problem is, that nobody at the site of the backup server has access to Veeam. I am just monitoring and operating Veeam. I am also not very satisfied with the statement of the support. They made is sound like it's an expected behavior, for me it looks more like a GUI bug, because of its randomness.

This happens since a week or two after the upgrade from 7 to 8, never noticed it before. I didn't found anything in the documentation or the knowledge base about that, maybe it's not that expected and someone can tell how to fix it, maybe the database is a bit messed up.

The support case ID is 00974090

Thank you,
Stephan

[veremin]

Do you mind answering several questions? Tape job is configured to export media once the job finishes, isn't it? Duplicated entity shows each time a medium is ejected and new one is inserted, right?

[Dima P.]

Hi Stephan,
Unfortunately, can’t read your case details - it’s in German. I have one question in addition: can you elaborate if the latest update 2 was installed after upgrading to v8? Thanks.

[Stephan23]

Do you mind answering several questions?

of course not.

Tape job is configured to export media once the job finishes, isn't it?

no, both media automation options (eject media and export media set) are disabled

Duplicated entity shows each time a medium is ejected and new one is inserted, right?

ejected physically at the server, yes. no veeam actions involved.

Unfortunately, can’t read your case details - it’s in German.

i knew i should have changes the preferred language

I have one question in addition: can you elaborate if the latest update 2 was installed after upgrading to v8? Thanks.

yes, update 2b.

[Stephan23]

am i missing the option to edit my post or is this not possible?

what if i just enable "eject media once the job finishes" ? i haven't done that, because i remember an issue that the tape got ejected before every backup was written to tape. but that applied to another infrastructure with a tape library and multiple tapes per media set.

[veremin]

It's a known issue with standalone drives that manually ejected tapes get duplicated. The issue is fixed in version 9, though.

And also should have a pretty simple workaround to let backup server eject tapes automatically, instead of ejecting them through drive's interface.

Thanks.

[Stephan23]

that's great news, thank you a lot.

i wonder why the support wasn't able to provide that information.

[veremin]

You're welcome. Let us know whether the proposed workaround fixes the problem for you. Thanks.

[Stephan23]

You're welcome. Let us know whether the proposed workaround fixes the problem for you. Thanks.

The workaround works. But I am foreseeing a much bigger problem with it, so that I can't use it, at least with the current configuration.

There a multiple Backup Jobs, each one does a full backup once per week on alternating days of the week. That way I can spread out the space requirements across the week, because I can only work with one tape. The problem is, that sometimes the tape gets full and the job continues in the morning when someone changes the tape. With the eject option enabled, that would cause the tape to be ejected after the job finishes in the morning.

[veremin]

And if I'm getting you right, the problem would be that a tape ejected on the morning could not be used then by a job starting at night?

[Stephan23]

Right.

edit (hey, i can edit my post):
I guess I'll wait till Veeam 9 and live with the duplication, until now no tape job failed because of this issue.

[veremin]

Do those duplicated drives have offline status? If so, you can probably write some simple PS script that would remove those entities from a backup console. If I'm not mistaken, the Remove-VBRTapeLibrary should accept drives, as well:

Code: Select all

$Drive = Get-VBRTapeDrive -name "Name of drive"
Remove-VBRTapeLibrary -Library $Drive

Be aware to test it first prior to implementing.

Thanks.

[hgrams]

Hi there,

Our single tape was exchanged by a new one via an external hardware provider. Now, Veeam shows both drives, the old one under Libraries and the new one as offline with all media pools assigned.
I cannot remove the old one and cannot assign the new one to the Libraries branch. I already rebooted the server and deactivate and activate the tape drive in Windows device manager.

What am I doing wrong? Or better what can I do to solve that?

Best Regards
Harald

[veremin]

Hi, Harald,

Does your issue look similar with the one discussed above? If so, try to remove stuck offline drives, using the referenced script, or ask support team for assistance.

Please, be aware that version 9 has this issue fixed.

Thanks.

[hgrams]

It doesn't work. I opened a case # 01007067

[hgrams]

The problem has been resolved by Veeam support. Here the solution and Veeam statement:

Now you need to do the following:

1. Right-click your ONLINE drive - Inventory. You will see no errors and it means the drive is working OK.
2. Create new media pool(s) on the ONLINE drive for your backup job.
3. Edit all your tape jobs to be addressed to the new pool(s) (on the "Media Pool" tab) 4. Then, when ALL the jobs are re-addressed (please check twice) you can delete the old drive.

It was confirmed officially by our R&D team that such behavior (two drives after replacing) is expected and changing database to delete the old one is strictly forbidden. The steps above is the only correct way.

[veremin]

Thank you for taking time and updating the topic with the solution proposed by support engineer.

According to the plan, version 9 should address this problem, so that you don't have to deal with workarounds any longer.

Thanks.

[pshute]

I've got the same problem, even after installing v9. I assume the duplicated drives were added over the last few weeks while we were still running v8 (which I installed a couple of days ago). I tried right clicking on the drive and running an inventory, but they're still there. All three of them list a different loaded media, and all say the state is "Loaded".

[pshute]

I've got the same problem, even after installing v9. I assume the duplicated drives were added over the last few weeks while we were still running v8 (which I installed a couple of days ago). I tried right clicking on the drive and running an inventory, but they're still there. All three of them list a different loaded media, and all say the state is "Loaded".

It was showing three drives, with tapes "weekly 1", "weekly 2" and "weekly 3" loaded. There is really only one drive, and at the time it had tape "weekly 3 in it".

I right clicked on the first duplicate (showing "weekly 1" loaded"), and told it to eject. Then I ran another inventory. The tape "weekly 3" ejected, and then it only showed one drive. But it said "weekly 2" was loaded (it was actually empty).

I reinserted "Weekly 3" and ran another inventory, and the display was now correct.

Then I manually ejected the tape and inserted "Weekly 4", and ran an inventory. Display is still correct.

Conclusion: The problem no longer occurs under v9, but this workaround is required to correct the display if it had already occurred under v8. Correct? Or would the problem have gone away by itself as I cycled through the tapes over the next month?

[Dima P.]

Conclusion: The problem no longer occurs under v9, but this workaround is required to correct the display if it had already occurred under v8. Correct? Or would the problem have gone away by itself as I cycled through the tapes over the next month?

This problem should be resolved in v9 indeed since media pools are no longer tied to a specific library or drive, thus device could be deleted from the UI without any worries. However, if you still have the described behavior I suggest opening a support case and let the support team investigate the debug logs. Thanks.

[Stephan23]

not sure if i should post this here, but it might be related.

as stated the issue was resolved for me with version 9. however there is now a new issue. every time a tape gets changed it doesn't show up in the gui. the old tape, that was in there before is still shown as online. only after a tape job has run (and only if is has written data) the online state of the drives gets correct.

i just created a new case: 01776483

[busitech]

This problem still happens in build 11.0.0.837 P20210507.

[Dima P.]

Hello busitech,

Please share the support case ID and I'll ask RnD folks to check your debug logs. Thank you!