instant recovery if primary VBR lost - how

[alesovodvojce]

We use VBR for complete backup of VMs on HQ.
Second copy of them goes regularly to the "DR site", via backup copy job.



The question is: when the HQ is lost, how can we recover the VMs in DR site?

As there will be only VBK+VIB files at the moment, and Veeam B&R config copy backup.
And in such disaster scenario, how can we prepare to be able to use Instant recovery only with data in DR site?

[ejenner]

You need a warm spare B&R server at the DR site. Alternatively you can change to or add a replication strategy. But that comes with its own questions so might not be suitable.

[DGrinev]

The best practice would be to have a second VBR instance at the DR site which will be responsible for replication from the HQ.
This way you can achieve a minimum downtime by using failover feature for critical VMs in a case of disaster.

Anyway, you should have a separate instance of VBR at the DR site that allows you to import backups and initiate instant recovery. Thanks!

[alesovodvojce]

Updated graphics below is reflecting your advices. That we should have warm spare VBR at DR site


Now we have two installation options:
a) to the VM that is hosting backup copy files already (pros: immediate import speeds, cons: can interfere with VBR infrastructure services versions deployed on the same machine by primary VBR)
b) to separate VM (pros: clean install, cons: slower ingestion speeds - backup data will be shared via SMB from another VM and at the time of ingestion copied again)

What do you (and best practice) think?
and @DGrinev, is there any other option than having remote DR site VBR to pull the backups from HQ site? Or our push architecture is wrong?

[DGrinev]

I would choose option "b" just to avoid any issues related to services, components, etc.
Your architecture for backups is good.
I was referring to the replication feature that can be implemented in parallel, it will give you an additional protection layer.

[alesovodvojce]

The b) is not working as expected. In order to do "Backups Import", I am forced to add the shared folder with backups in the infrastructure. So, add whole server into the infrastructure, and mess with his already installed components.

Variant a) now seems as a much easier option. I will test it and let you know

[alesovodvojce]

We have been adviced to change to topology completely.
Instead of orchestrating Veeam VBR from HQ site, to orchestrate it from DR site and use HQ only as backup proxy.

This way when any accident occurs, we will be able to access backups immediately.
But how about minor accidents when DR site goes offline but HQ does not, can it be solved by installing another VBR management console to HQ?

[DGrinev]

I would not expect component issues if both instances will be kept on the same product version.
The topology is similar to the one that was proposed by me in the first reply. Thanks!