Quick question on CBT

[sporgg]

Hi all

I am currently using Cluster Shared Volumes on a Server 2012 Hyper-v Cluster. It is not going well at all. Thinking of getting rid of the CSV volumes and going back to normal volumes. One Per VM. What I want to know is that if I do this will Veeams Block Level tracking still work ?

Thanks for the replies.

[Vitaliy S.]

Hi Gary, yes, it will work.

[sporgg]

Hi Gary, yes, it will work.

Thank you for the info. I have had enough of csv volumes for a lifetime. Now I can start migrating away from them and know I can get a Backup working.

Thanks again

[Vitaliy S.]

Out of curiosity, what kind of issues did you have with CSVs?

[sporgg]

Out of curiosity, what kind of issues did you have with CSVs?

Bit of a long story there,

Main problem is a memory leak issue that has been there since day one. Second issue is that during backups the csv volumes go offline and crash the servers. I have been working very close with Veeam support and must say they have tried every trick in the book to get it working but I have proven that the problem is not them and is an issue with CSV. Together we have tried several Microsoft hot fixes that should all work in fixing the issue but for me the memory leak just wont go away. Veeam support have not given up and are still trying to find a fix for it but I just cant wait any longer. So have decided to move away from the CSV. Cant thank Veeam support enough and would recommend them to anyone having problems. But CSV is just not ready yet for what I need.

Any more questions feel free to ask.

Regards

Gary

[Vitaliy S.]

Just talked to our QC guy and has confirmed that they've seen similar issues with CSVs. Also wanted to clarify my previous answer... are you going to use clustered disks by any chance?

[sporgg]

Just talked to our QC guy and has confirmed that they've seen similar issues with CSVs. Also wanted to clarify my previous answer... are you going to use clustered disks by any chance?

Hi

Yes looking at Clustered disks at the moment. Is there another better way ?

For more info I am also looking at how I may be able to use pass-through disks

Regards

Gary

[Vitaliy S.]

Another way would be to go back to using CSVs Ok, since you're planning to use clustered disks, change block tracking for these disks will only be supported in Veeam B&R v7, that should be available soon.

[sporgg]

Another way would be to go back to using CSVs Ok, since you're planning to use clustered disks, change block tracking for these disks will only be supported in Veeam B&R v7, that should be available soon.

Ah ok, So apart from CSV what volume would CBT work on ? Pass-through disks ?

[veremin]

Actually, Pass-through virtual disks, as well as, disks that are connected via in-guest iSCSI initiator are not supported and are automatically skipped from processing during backup/replication job.

Thanks.

[sporgg]

Hmm

Ok, anyone got an idea what I can do. The hyper-v runs a production site that works 24x7. So having servers crash during backups is a real big problem. Also the performance issues with the memory leak again cause issues. I need a way to run these vm's and make backups without causing issues to the production systems. CBT is a big thing for me on one of our servers as it is just over 1TB and growing. But on the others I could live without it. But yeah, Anyone got any thoughts ?

Regards

Gary

[collinp]

I've had similar issues. We have a very large Hyper-v environment with over 125 VM's. I have come to the same conclusion as you that the weak link in Hyper-v are CSV's. It seems like there are a lot of people having these issues while using many different backup products. I'm curious as to what hardware you are running so we can compare. Standalone vs. Blade Architecture, SATA/SAS drives, Storage Vendor, Software/Hardware VSS, etc. Also, what errors are you seeing in Failover Cluster Manager during the outages 1135, 1146, 1230?