03-17-2021 01:23 AM
Hi,
we changed a simple full backup-to-disk-to-tape to a full/diff schedule because of time reasons.
Since then, GRT warnings apear.
BE version 20.6
Environment: Server 2019 Hyper-V Cluster
GRT: File and AD
Before the change, the job ran without warnings.
Now, the new job (full/diff) has a warning regarding the copy to tape/GRT.
Drive F: of that server suddenly can not be *copied* to tape without a GRT warning.
Nothing changed on that server. Restart also was done.
Does anyone have a clue, what could be the issue?
Best regards,
Daniel
03-21-2021 11:39 AM
03-23-2021 07:43 AM
Hello Lothar,
thanks for your ideas. None is the issue.
But I'm a step further.
While reading the warning again and again, I noticed, that we do not have an issue with the parition "f:" on the backuped server but
with the local backup-to-disk drive on the backup-server.
The warning mentions, that "f" is not grt-capable. "f" is meant to be the f.vhdx of that server under the x:\BEDATA\IMGxxxx folder.
Interesting is, that the issue only occurs on that b2d folder/item while doing the diff-backup to disk.
Checkdisk showed no erros on the local disk-space.
I'm now testing the change from instant-grt to regular-grt. So far, the first job went through without warnings. Curios.
I also disabled testwise the antivirus. Logs (Windows, BE, antivirus) where no help, except showing a warning regarding:
Backup-Server B2D space:
System - Ntfs, ID 55
In der Dateisystemstruktur auf Volume "E:\BEData\IMG004272\MP00\F" wurde eine Beschädigung erkannt. Die genaue Art der Beschädigung ist unbekannt. Die Dateisystemstrukturen müssen online überprüft werden.
This must be some temporary staging-file/folder. It just occurs while doing the diff-backup.
Best regards,
Daniel
03-23-2021 08:39 AM
Hello Daniel
"E:\BEData\IMG004272\MP00\F" is the Folder where the post processes of GRT mounts the vhdx-File to do the instant GRT and regular GRT infomation collection.
If there was an crash sometimes the folder retains and other mounts fails until a new full backup is done.
If the folder is empty you can delete it normaly over Windows Explorer (sometimes after a reboot).
Please remind, all the corresponding Full and diff backups of the past days have to be accessable on disk while GRT is collecting the file or volume infos. If not then the GRT fails in my mind.
If media lifecycle deleted them this could be the problem.
Remind: In Test scenarios the data is not old enough, so media lifecycle retains the media on testing.
Kind Regards
03-24-2021 12:57 AM
Hello Lothar,
thanks again for taking your time.
I already had a look after that folders and orphaned data, but nothing indicated some kind of crash.
Also the "rotation" full/diff is correctly working.
It's weird.
The solution in our case was to switch from Instant-GRT to Regular-GRT.
Don't ask me why, but it went through without warnings. I just could find that out through under exclusion search.
Summarized:
- Daily full-backup (B2D2T) of several Hyper-V VMs with Instant-GRT worked fine
- Changed to Full/Diff in a seven day schedule with same setting -> warnings regarding the mentioned
B2D folder on the backup-server occured
- Changed from Instant-GRT to Regular-GRT: Solution, no more warnings in the diff copy jobs
If you or anyone els has a clue, what could be the issue, I would be very interested.
Best regards,
Daniel