Creating A snapshot issue

[ZakMc]

I have set up Veeam with Vmware Appliance, and I understand it has to take a snapshot each time it backs up, but the issue seems it takes a snapshot with virtual memory which freezes each live VM for about 5-10 minutes. Each VM has a different storage array and is not on the same array and generally on SSD's. I know you cannot turn off snapshots for this process as read in previous posts, but is there something to stop the complete freeze of these VM's during the snapshot since as I said I feel it was using a snapshot with virtual memory as this is the produced behavior I always get with that checked on doing a manual snapshot (which is why I don't check this option). I am Running Veeam Backup and Replication Communoity Edition (latest edition) and Vmware ESXI 6.0 (yes I know, but I have machines I need to use that will only work nice on 6.

[Gostev]

Actually no, Veeam does not take snapshots with virtual memory in principle.

[ZakMc]

Would you know why it takes a long time to create a snapshot then? A local snapshot takes 2 seconds, via veeam request takes 5 minutes.

[Mildur]

Are you using a NFSv3 Datastore and do you have installed veeam in a vm (Veeam with Vmware Appliance)?
If yes, check this article. There is a known issue with NFS v3 and Hot Add Backup method. Could be possible, that you face that in your environment.

https://kb.vmware.com/s/article/2010953

[ZakMc]

Mildur thanks for the response.... I have it set to an External HDD off a NAS system, the HDD is formatted in NTFS and I believe it's a SMB type, I tried to use the NFS mount via windows and veeam could not find it when I populated the drives in setup.. Not sure if this could also be an issue, it is also hosted on a windows machine as well.

[Mildur]

My question was, how is this storage (HDD connected to the NAS) connected to your vsphere environment. The storage, which your vms are using.
Is it NFS or iSCSI? What protocol are you using?

[ZakMc]

Gotcha, I actually don't have this storage mounted via vSphere at all... I just have it connected via Veeam via the software, this nas/hdd does not even hit vSphere directly.
But the VM's itself are on an iscsi lun via the NAS.....

Edit: I again read your question wrong, I am using iscsi for the VMs itself.

[PetrM]

Hello,

Did you try to create a snapshot with VMware Tools Quiescence in vSphere Client and did you check whether the corresponding option is enabled in job settings? Also, you may consider Veeam Agents managed by Veeam B&R to protect your workloads as agents work inside the guest OS and there is no interaction with vSphere, therefore you won't need to troubleshoot the snapshot issue which happens at the level of hypervisor.

Thanks!

[ZakMc]

PetrM,

So I found two interesting things out, If I had any pre-existing snapshots (which I know documentation says to delete any local snapshots) then this 3-5 hang would occur, but if I left a native snapshot and disabled the Changed block tracking, then I would not get the hang at all. Furthermore, obviously, if I deleted all vSphere snapshots for that VM and left the CBT on it would then also not hang...

[PetrM]

Hello,

In this case, the main goal is to make sure that you don't have snapshots on VMs you're going to protect and you are good to go!

Thanks!