Our current VM farm is backed to a local Linux Immutable Repo and then BCJ to another Linux Immutable Repo in DR site. Both site are L2 connected with metro cluster formed on storage level.
VBR server is installed on VM farm.
What if: the production site is down, my VBR server is rebooted on DR site; while my primary backup repo and production VM farm is gone. During this period, I will leverage of resources on DR site and backup is still required before failback to production.
I have to stop all existing backup policy and create new backup policy because backup path is changed?
The initial backup during that moment will be incremental or full? (wondering because all backup file are there, but there are no LUN level replication before)
Veeam jobs are working based on the VM UUIDs from the hypervisor. I assume that in this case (depending on your hypervisor system and the DR capabilities) the VMs and so the UUIDs stay the same when it comes to a failover so from my perspective you can simply use the existing and working backup jobs. It may be that CBT won’t work but basically you should expect incremental backups as usual.
For failback this is the same scenario since the UUID will hopefully also remain the same.
Please not that this is for DR / HA scenarios, e.g. a vSphere cluster split across two sides. In this case you don’t have to worry much about your UUIDs and you backup chains.
In case that you don’t expect the UUIDs to remain same you might have to create a new job or expect the existing job to create a new chain starting with a full backup.
Veeam jobs are working based on the VM UUIDs from the hypervisor. I assume that in this case (depending on your hypervisor system and the DR capabilities) the VMs and so the UUIDs stay the same when it comes to a failover so from my perspective you can simply use the existing and working backup jobs. It may be that CBT won’t work but basically you should expect incremental backups as usual.
For failback this is the same scenario since the UUID will hopefully also remain the same.
Please not that this is for DR / HA scenarios, e.g. a vSphere cluster split across two sides. In this case you don’t have to worry much about your UUIDs and you backup chains.
In case that you don’t expect the UUIDs to remain same you might have to create a new job or expect the existing job to create a new chain starting with a full backup.
Have you looked at using Veeam Replication? Do you have hardware at your DR site to use as a ‘cold site’?...with the only thing running is a 2nd VBR server?
The above is what I do for DR. I have 2 VBR servers → DC1 VBR for main production local backups and a DR-side VBR server which I solely use to do Replications from DC1 → DR site. If my DC1 site goes down...no big deal; well...it would be a big deal 😉 ...but I’m just saying I have everything I then need to get up and going fairly quickly at my DR site cuz 1. a complete and “ready” replicated copy of each of my prod VMs are already on Clustered VMW Hosts, and 2. I can get my DR process going readily cuz my 2nd VBR server is up and going already.
Anyway...just something to consider. Backup Copies are good..you have an offsite copy of your Prod, but in the event of a DC1 outage, using Veeam Replication is the best way to go imo….if you have the infrastructure to do it.
Best practice is run Veeam B&R server at the DR site. If your main site goes down, you can restore and run replicas from DR site. If DR site goes down, things stay running until you get it back up. You could have a cold standby at the main site just make sure you are getting your config backup over to the other site also.
Running an additional Veeam server specifically for replication and keeping at the DR site is an option too.
UUID, and DNS. Veeam should have no issue when you run from the other site. I always have manual hosts files on my Veeam servers, repos, proxies etc. because if DNS fails, you are going to have issues. (been there before)
If you have a repo at both sites, hopefully you have a proxy at both sites too, just move your Veeam Server VM to the DR site and you are good to go.
Have you looked at using Veeam Replication? Do you have hardware at your DR site to use as a ‘cold site’?...with the only thing running is a 2nd VBR server?
The above is what I do for DR. I have 2 VBR servers → DC1 VBR for main production local backups and a DR-side VBR server which I solely use to do Replications from DC1 → DR site. If my DC1 site goes down...no big deal; well...it would be a big deal 😉 ...but I’m just saying I have everything I then need to get up and going fairly quickly at my DR site cuz 1. a complete and “ready” replicated copy of each of my prod VMs are already on Clustered VMW Hosts, and 2. I can get my DR process going readily cuz my 2nd VBR server is up and going already.
Anyway...just something to consider. Backup Copies are good..you have an offsite copy of your Prod, but in the event of a DC1 outage, using Veeam Replication is the best way to go imo….if you have the infrastructure to do it.
