online backup and offline backup

[marieelieg]

Hello everyone.

I have read the difference between online and offline backups in the veeam guide for hyper-v.
What I want to know is how to configure them in Veeam 9.

In the Guest processing step in the backup job creation wizard, we can either enable application-aware or disable it.
Does this mean if I enable application-aware, it is an online backup and if I disable application-aware, it is an offline backup?

Or do I have to do anything else?

Thank you in advance:D

[Mike Resseler]

Hi,

If you use application-aware processing (and it is possible depending on the workload) then it will be an online backup and you will get an application-aware backup which is preferred to have consistency for your data. If AAIP cannot be used, you can go in the advanced backup settings - hyper-v and choose Enable Hyper-V guest quiescence. In this case, the following will happen:
* If a VM meets all requirements for online backup, Veeam Backup & Replication will use the online backup method.
* If a VM does not meet requirements for online backup, Veeam Backup & Replication will use the offline backup method.

So basically, even if you choose Hyper-V guest quiescence, you might be able to have an online backup, but it could be an offline backup (meaning the VM will go into a suspended state for a very short time)

Last thing, if you don't want the VM to become suspended, you can select Take crash consistent backup instead of suspending VM which means we will take a crash consistent backup and not suspend the VM.

Hope it helps

Mike

[marieelieg]

I see so even if application-aware is or is not enabled, if the VM meets all the requirements an online backup will occur.

[Mike Resseler]

Correct. But this is a good thing, or do you have specific reasons for not wanting to take an offline backup?

[marieelieg]

No no I just want to know the difference (if there was any) in configuring both backup methods. That's all:D

[Mike Resseler]

Cool, hope I could help then

[marieelieg]

Yes you did thanks:D

[elwood]

Is it feasible that an Application Aware could fall back on an offline backup? We're seeing results consistent with an offline / saved state despite using app aware.
The symptom is that Exchange is pauses for just a few seconds, twice[ during the backup (freeze / unfreeze? )

We're scouring logs to try to find what kind of backup it's actually performing and I see things like 'VSS Snapshot mode : 'Enchild' (Googling reveals nothing - what's this? ) and "Snapshot CTP unavailable", is that good / bad?

Is there a real easy way to see with absolute certainty that the App Aware backup is working correctly and in Online mode?

[Mike Resseler]

Hi Eliot,

Are your Hyper-V integration services up to date?

[elwood]

Yes, bang up to date on both host and guest.
We have the issue occurring on a clean-built standalone (non cluster) host as well as a clean build (and patched to the May 17 rollup) cluster.
Other than the two blips we distinctly see during the backup, the other clue is the fact that the uptime counter gets reset on HyperV, which according to the KB below suggests that it's not running an online backup.
https://www.veeam.com/kb1896

[Mike Resseler]

This is my list of items you need to check to make sure online backup is being performed:
* The latest Integration Services must be installed and backup (volume snapshot) within those services must be running
* The guest OS inside the VM must support VSS
* Is the Snapshot File Location for the VM set to be the same volume in the host operating system as the VHD files for the VM?
* Are all volumes in the child VM basic disks and are there no dynamic disks?
* Are all disks in the child VM use a file system that supports snapshots (for example, NTFS, ReFS)?
* Following services must be running in the VM: COM+ Event System/ Distributed Transaction Coordinator/ Remote Procedure Call (RPC)/ System Event Notification/ Volume Shadow Copy. (Some are set to start automatically, others automatically (Delayed start). The COM+ System Application/ Microsoft Software Shadow Copy Provider services must at least be set to Manual
* Check inside the guest for event ID 13 with source vmicvss. If that exists, something is wrong

If all checks out OK, please log a support call and post the case ID + follow-up after support investigation here
Thanks
Mike

[elwood]

Hi Mike,

We've actually concluded that there is a problem within the Hyper-V VSS mechanism which causes this disconnect. One of your support engineers concurred with this view after we ran some Diskshadow backups independently of Veeam and confirmed that the issue is still present. .
We know that Hyper-V presents a brief vNIC re-connect, followed by a check of system time (event IDs 14 and 1) after a standard live-migration (normal behaviour). As a side effect, the uptime counter in Hyper-V manager is reset. What we cannot explain is why this same set of events happens during a quiesced or app aware backup with VSS - and at the exact moment of the vNIC reconnect we see a brief pause on the VM. Research suggests that behaviour is seen in a saved state backup, which suggests that one of the pre-requisites for an application aware backup hasn't been met. However we've exhausted all routes to isolate an obvious problem. All of the pre-reqs you list above are met.
It seems that users of rival backup products have experienced similar things on Commvault, Acronis and others too, so it certainly looks to us like an MS problem.

We're working with Microsoft Support at the moment to try and resolve this. I'll let you know if we discover anything useful.

[Mike Resseler]

Hi Eliot,

Letting us know if something is discovered would be very helpful and we would appreciate it a lot!

Many thanks
Mike

[elwood]

Hi Mike,

We were hoping that in the changes Microsoft have made to VSS for 2016, and the new integration mechanisms through Production Checkpoints that we'd see some change in this behaviour. Alas not the case.
We upgraded a cluster to Functional level 9, some clean build 2016 hosts and a 2012 based VM testlab. VMs upgraded to version 8.0 and ran all updates (for Integration Services updates now being through Win Updates). We see the exact same set of results, the NIC reconnection (event ID 12) followed by a time reset (Event ID 1) and at that exact moment in time the Exchange is frozen for anything up to a few seconds.
This looks to be something inherent to Hyper-V and specifically the interaction between the host and guest OS/Application. I suspect that it's actually fairly common but only customers running their backups during working hours are likely to notice it.
We still have our case open with MS so will provide any update if we find a solution. Unfortunately 2016 doesn't seem to be it!