How to Start Over with VDR?

[ph0bia]

VDR screwed up *again*, vaults all messed up, all backups failing, no repair actions help.  Had to re-format the vaults.   Does anyone actually count on this product for data protection?  We  use it as a nice-to-have extra layer only, to do otherwise would be total  lunacy.  I've never used a worse piece of backup software and I go all  the way back to Fastback and 5.25" floppy disks...

Anyway all ranting aside, is there any way after re-formatting the vault to make VDR start over with brand new backups of the VMs in the existing jobs?

The existing backup jobs just keep failing saying they can't find the directories on the vault, throwing an -1102 error.

I tried duplicating the backup job to a new job, but same effect when it runs, it is trying to look for all the old stuff which is gone.

I don't want to have to go through the whole process of rebuilding these stupid backups yet again, please tell me there is some way that I can tell it that I just want it to start again from scratch.

[cag201110141]

I feel you pain. I gave up on VDR 2.x after 11 lost datastores.  I am now using veeam backup and replication. But what I did find that helps is having a high ratio of VM's to backup to the VDR vault.   I was backing up 520GB to a 380GB vault which works for about 30-45 days. Then all hell breaks loose.  What is your ratio.  There are some setting that could be used to fix 2241 errors but this only helps for a while.  I have had a open case since June I think ?. Dont remember its been so long. I am still waiting for a resolution. What I did do is create 2 vaults so when 1 went down I could use the other. As for the jobs schedules I was able to continue using the existing jobs only after re-editiing and pointing to the new destination. Have you tried to re-edit the jobs ?

[phykell]

What's your current configuration? Are you using NFS, CIFS, etc.? How many VMs are you backing up and what type of VMs are they?

It seems to me that before trying to rebuild all your backups again it might be worth looking at the overall architecture because if there's one thing I'm sure of with VDR, it's not very tolerant of potential issues such as slow storage targets for the destinations. I find it works just fine with a good NFS NAS and backing up a small number of VMs to a relatively large destination but then it is supposedly an entry-level product for small deployments and perhaps we users expect too much of it!

I'm looking at replacing with Veeam as our requirements grow - if you consider the amount of time you spend on getting VDR to work how you *need* it to, the cost of a third-party, dedicated product is less of an issue in my opinion.

Sorry that's not really a definitive answer for you, but if you can show your current architecture that might be a good start to solving your problems long-term.

[ph0bia]

Thanks everyone for your replies. The vaults are FC-SAS on NetApp, they are quite bored.

I am protecting around 180 virtual servers, using 3 VDR servers and three separate LUNs on the NetApps.

Have heard great things about Veeam, going to check it out

[Jim4649]

Check out AppAssure, too. Cheaper than Veeam and just as good.

http://www.appassure.com/server-backup-and-replication/vmware-backup/

[phykell]

Sounds like there's nothing wrong with the hardware you're using though 180 server VMs shared amongst just three VDR appliances is probably too much for VDR. Personally, I'd recommend no more than a few large VMs or up to 20 relatively small VMs per VDR appliance. I'd even suggest having one VDR per physical host assuming you have more than three.

Regarding AppAssure, we did consider them ourselves but there was no support for Linux VMs unfortunately. AppAssure have just been acquired by DELL so that bodes well for them in the backup space. Hopefully, they'll be able to offer Linux support in the very near future.