How to comply the AVAILABILITY for the Always-on Enterprise

[foggy]

Nothing to add to what Fedor has stated above - we cannot comment on the future release plans, but all the votes and reasoning here are taken into account. Thanks!

[rdbastedo]

Now that V10 of Veeam B&R is released can someone comment? I have been following this and am hopeful.
We have primary and secondary sites, synced 70TB repositories on a dual 10G link.
Veeam B&R running at the secondary site with the primary backup repository there.
Mirrored repository at the primary DC.
Our VBR is on a Hyper-V VM that is the only VM on the host.
I'm aware we could do a VM copy to the primary DC however having some kind of native HA within VBR would be amazing.
The primary site has an identical server just waiting to be installed as the HA VBR machine.
Currently we do a backup of our configuration to the cloud and have our encryption keys backed up as well.
I would love to get away from the plan of spinning up a new VBR instance and then installing the backed up config being our only choice.

[foggy]

Hi Rick, v10 doesn't change anything in this regard. Options discussed in this thread are still the only way to achieve high availability of the backup server. Thanks!

[rdbastedo]

Thanks Foggy, I'll go ahead with one of those options.
Good to know this before putting in the effort.

[fawwaz93]

Hi All,

As the subject suggests, can we actually create a cluster of a VBR where if 1 of the VBR server crashes,the other VBR server can run the existing jobs as usual without any downtime? (Something similar to HA) This concern comes from one of our clients which explains to us the current VBR setup which we install in a physical server, might crash without any signs since its an old server.

If there are no cluster options on the VBR level, any thoughts/opinions on how we should handle this challenge in the near future?

Thanks.

[wishr]

Hi Fawwaz93,

Currently, there is no native solution for that, however, you may use one of the aforementioned workarounds.

Thanks

[micoolpaul]

Hi,

Wasn't sure which of these topics to contribute to, but I wanted to throw a +1 into the feature request for some sort of high availability/survivability of Veeam.

Lets be honest, we all know that the big wish is a fully fault tolerant VBR, but that takes a lot of work to get feature parity with even some of the more "basic" features, let alone tasks such as CDP that are continuously working.

I would love to see a multi-site aware VBR product as an evolution towards a fully fault tolerant VBR solution.

What do I mean by this?

Presently, to have a low RTO in the event of a DR, you tend to install VBR onto a DR site, this unfortunately has some drawbacks, if you have a connectivity break between your production & DR sites, you don't have functioning backups, leading to RPO violations.

I'd like to see the ability to deploy multiple VBR instances, that are aware of the others. Job orchestration is handled by the VBR at its local site, but DR functionality is syncronised to the other sites. For example, failover plans. This way you get the benefits of local job processing, but can still trigger DR capabilities during an actual DR event.

[HannesK]

Hello,
I see several requests in that post. It's very unlikely, that we built an active-active grid of VBR servers that can deal with network outages in the short term. But we want to improve the situation at least to have some kind of active / passive scenario that is useful for planned maintenance or unplanned outages (no timeline available)

CDP continues working even without backup server with some limitations.

Component reconfiguration requires Veeam CDP Coordinator Service to be working. If the coordinator goes down, existing CDP policies still work, create and remove short-term restore points. Long-term restore points are not created and removed because it is the coordinator who manages them. However, if any of the components goes out of service, for example, VMware CDP proxy goes offline or VMware vSphere vMotion changes the infrastructure, CDP policies start failing until Veeam CDP Coordinator Service is repaired.

For failover plans: I think Veeam Disaster Recovery Orchestrator would be a way to solve that.

Best regards,
Hannes

[vmtech123]

Even active / passive would be handy. My thought is in a REAL disaster where my main site goes BOOM. I have SRM to flip the switch and fail over to DR for critical stuff currently (Would love to get this in Veeam), but the rest of the backups I need to restore. Not having to install VBR, import config, add proxys etc would be a game changer. That would save a significant amount of time.

[micoolpaul]

Morning Hannes, sorry I missed your reply, I appreciate we’re blurring topics now but I don’t see VDRO fixing this, as VDRO still only has single instance support, and each VBR can only be managed by a single VDRO instance. Did my example give you an idea of a specific topology that would be appropriate?

[HannesK]

Hello Michael,
if multiple orchestrator servers are no option, then I suggest to discuss that topic separately in the Orchestrator forums.

A multi-master VBR grid, that works even when links are down is something, that will probably never happen (I know, never say never )

Best regards,
Hannes

[ledvanb]

Dear Colleagues,


I would like to know if anybody know or has found an alternative for Veeam Backup Server HA capability? i want my backup server to be Highly available with zero downtime if under any circumstance the Backup server went down.

regards ,
Ledvan.

[ChriFue]

If your VBR server is a virtual machine and on VMware, you could use VMware Fault Tolerance.
Only possible if the VBR server is in usage for management only.

You can use vSphere Fault Tolerance (FT) for most mission critical virtual machines. FT provides continuous availability for such a virtual machine by creating and maintaining another VM that is identical and continuously available to replace it in the event of a failover situation.

https://docs.vmware.com/en/VMware-vSphe ... 2ED21.html

Yep, Enterprise Plus and +10G Ethernet needed ...

[ledvanb]

thanks @chrifu for relying on such short notice, the server is an all in one Physical solution which holds all the roles(VBR+proxy+gateway). is there any other workaround?

[ledvanb]

anyone can help with this?

[Mildur]

Hi Mehrdad

I moved your post to this topic.
Currently there is no native support for HA in VBR. Please see the previous answers in this topic on how to achieve lower downtimes.

Thanks
Fabian

[ledvanb]

Hi Fabian,

Thanks, will do so.

[ledvanb]

anyone has tried windows fail-over clustering + sql always-on availability Groups to reach some sort of HA? if so, did you face any issue ?

lets say we implement it ,if one of the nodes(main backup server) fail and backup start on the another node(fail-over), will it start from scratch or resume the job? i would appreciate if anyone share his/her experience with configuring Veeam+WFCS+always on availability?

regards,
Mehrdad.

[HannesK]

Hello,
SQL alwayson works (documented in the user guide)

Putting the backup server on a failover cluster with a shared disk... not sure how this could could work. Things like registry keys are on C:\. Also the second backup server could not decrypt any password from the configuration database, because the passwords are encrypted with the machine key.

Best regards,
Hannes