Veeam Server DR

[bezza]

I have a veeam server in my production site sending replica's of all my virtual machines to some ESX servers in my DR site. Everything is working perfectly, i can failover VM's etc and all is good. However if i was to suffer a complete site failure at my primary site, how do i fail over to my DR vm's as my veeam server would have been taken out when my production site went down?

My question is, does veeam provide a solution for replicating my 'veeam' server to another 'veeam' server in my DR site?

Thanks,
Aaron.

[Alexey D.]

Hello Aaron,

Please refer to the existing topics with answers to your questions:
How to failover: DR Site Replication
How to protect Veeam B&R server: What is the best way of protecting the Veeam Backup server?

[nickf]

You won't need a Veeam server just to bring up the replicas (just follow the process Alexey links to), we've recently tested this (16 VM replica restore) and it worked great.

[GraPghaST]

[Merged into existing discussion]

We have an environment of 3 ESX4 hosts on a Equallogic SAN. We replicate VM's with Veeam B&R to a Disaster Recovery site, with one ESXi host and local storage datastore. Now we want to create a complete DR-setup. My main question is: How to make it possible that, in the event of a disaster, I can start the VM replica's on that remote ESXi host?

Plan is as follows, please let me know if I make a mistake somewhere.

- Every night all VM's get replicated to the Disaster Site
- Veeam runs on a local machine, so I made a job that exports the MSSQL Express database to a .BAK file, and copies it every night to the Remote Site
- On the remote Site ESXi host, I created a new VM with a clean veeam install. Nothing connected.
- On the Remote Site ESXi host, I recreated the Virtual network present in the Production Site
- On the Remote Site, I installed a vCenter VM and connected the ESXi host to it
- In case of disaster, I import the replicated Veeam SQL database into the Veeam installation at the Remote Site. After restarting Veeam services, I am then able to start up replica VM's in case of Disaster at any restore point present. When new hardware arrives, I am then able to Storage vMotion the VM's back to the new production environment.

is it also possible to import the database into the Remote Site Veeam installation, to check if I am able to "see" and startup VM's at the Remote Site without disturbing the current replica jobs & data? Or does the imported database mess up my replica VM's? Is vCenter necessary on the remote Site for Storage vMotion back to production? What is the best way to fallback when new hardware has arrived at the production site? Connect the ESXi host to the new vCenter farm and then Storage vMotion? Maybe attach the ESXi host to the existing vCenter cluster? It gets replicated every night, so theoretically I can start op the replica vCenter VM and connect to the ESXi host. Or won't that work?

Many thanks for all help and answers!

[Alexey D.]

Hello Robbert,

Please follow the abovementioned links, they should cover all your questions. Thanks!

[GraPghaST]

You won't need a Veeam server just to bring up the replicas (just follow the process Alexey links to), we've recently tested this (16 VM replica restore) and it worked great.

Yeah, but then you can't choose restore points in time. And, what is the replica job was only halfway when it got aborted or when the Disaster struck? Then you have an inconsistent replica, which can't be started up manually. A full resync is the necessary after that, which won't be fun when performed over WAN connections.

[Vitaliy S.]

In this case, you should be protecting SQL configuration database of Veeam Backup server (as you do it right now judging by your other post). Another way would be to use "pull" replication method (install Veeam Backup server on the target site).