VDR 1.2 shows huge Non-dedublicated Size

[vesajyl]

Hi,

today VMware Data Recovery crashed while it was doing integrity check. As a result, backup destination details now show Non-dedublicated Size: 126 TB. I'm pretty sure that this isn't right, Dedublicated size is showing 283 GB (it is probably correct). I have about 25 vm's to backup.

Should I be worried? VDR version is 1.2. As a backup destination I use QNAP NAS with CIFS share.

Integrity check crash was probably caused by me: automatic integrity check had been going on over 24 hours and thought if I could make it faster by deleting some test machines backup jobs. After I clicked test machine backup job and selected Delete, for some reason backup appliance crashed. After it restarted and made recatalog, result was quite big non-dedublicated size.

-Vesa

[Paul11]

I also have Non-dedublicated Sizes which I can not believe that they are true. But I don't worry about that, as long as the Integrity Check finshes without errors. Until now there are no problems also these values are unbelievable.

[AndreTheGiant]

Welcome to the community.

I suggest to do not use CIFS, but instead export a disk with iSCSI from your storage.

In this way you can use it as a VMFS datastore and can attach a vmdk to VDR.

Andre

[vesajyl]

Ok, I will try that. That QNAP NAS supports also iSCSI, so i define there one (or two?) about 1000 gb iSCSI disk and map them to our cluster and from there to VDR appliance.

-Vesa

[AndreTheGiant]

VDR documentation suggest to have destination smaller than 500 GB.

Create 2 different destination instead of one large.

Andre

[billylucas]

Question though... Why is it that VDR reports huge amounts of deduplicated data? I have a 1TB VMDK mounted (granted I should probably be using 2 500GB) but none the less, reports that I have 160GB deduped and 4.22 TB that isn't. What is this non-deduplicated size? Is that what it would be if I had all generations of data backed up with no deduplication?

[billylucas]

anything?

[vesajyl]

Hi,

I thought now would be a good time to update this thread and tell you newest event's with VDR.

As suggested, I added new iSCSI disk (size 493 GB) to vSphere and added it in VDR in raw format. In VDR appliance I formatted it and it shows as /SCSI-0:1.

Problem was that all my backups were in CIFS share and I didn't want to delete them. So I configured backups go to "Local Volume" waited couple days, marked all the CIFS share backups to delete and tried to run integrity check. As usual it runned couple days and then VDR crashed.

As a result Non-dedublicated size of local disk went from 1,5 TB -> 36.0 TB. And old backups are still in the restore list.

Question is, can I just unmount CIFS share? What happens all my backups when they are separated in two volumes (CIFS and "Local Volume") ?

[Paul11]

I have a similar configuration. Yes, you can unmount the CIFS-Share. Than you will lose only the backups of this location. In the Restore-Tab you get a merged list of all the backups of the two locations. If you unmount one location, you only lose the backups of this location. I'm using two RAW devices for backups, one for Monday, Wednesday and Friday, the second for Tuesday, Thursday and Saturday. So it dosent matter, when the Integrity Check needs more than 24 hours, as long as it finishes bevor 48 hours. And if one datastore gets corrupt, I have a second one to do a restore.

[vesajyl]

Ok, I will try that. I really really hope that VMware would improve Data Recovery and fast. It's so annoying when you can't trust if the backups are working or not.