Backup & Replication 6.5 - DR Architecture Design question

[wkuypers]

We are using Veeam Backup & Replication v6.5 (currently only for Backup purposes) and would like to start using the Replication feature to replicate to our DR site.

Setup is as follows:

Site A:
3 Hosts, 2 SAN’s (15TB) connected via iSCSI running VMWare (40 VM’s)
1 Physical Server (Running VEEAM Backup & Replication) with 2 SAN’s (20TB) connected to the same iSCSI network (Direct SAN Access) using Reversed Incremental Backups with a 30 day retention period.

Site B: (DR site over WAN link)
5 Hosts, 4 SAN’s (40TB) connected via Ethernet
1 Physical Server (proposed to run VEEAM Backup & Replication)

The scenario is as follows:
We will be using the replication part of the software to replicate all VM’s from site A to site B (the idea is to have 2 replica copies per VM)
Now, should site A fail completely, we have a replica of each VM at Site B, however the Backup server is situated at Site A and therefore we can’t use it to fail over to the replica, but instead we will have to manually power on each VM at the DR site (without the backup server knowing about it which in turn will cause a problem when Site A comes back on line and we need to replicate the changes back to Site A and make this site live - Fail back)

Possible Solution:
Option A:
I have installed and configured another backup server at Site C and tried to manage all backups from this server so that in the event of Site A or Site B going down, my backup server won’t be affected and I can do the fail over and the fail back. However, in doing this, the backup server creates an enormous amount of traffic on the WAN link (even though I have a proxy server at Site A). The traffic is so much that this solution does not help me at all. Unless I have configured something wrong in this setup.

Option B:
Another thought would be to run the Backup Server from Site B and manage all my jobs from there, so If Site A went down, I would still be able to manage the failover and failback jobs via the software. However I then still have the situation where each backup job (now managed from Site B) will create a large number of traffic on this links which is dedicated for replication – even if I have backup proxies configured?

So my question is:
What is the best design for this scenario, (taking into account I only have licenses for the 3 hosts in Site A) which will enable the Backup Server to always be available to manage the entire DR process?

[Vitaliy S.]

Hi Willem,

I would recommend to install a secondary Veeam B&R server on site B to manage replication jobs only. I don't believe there should be any difference in terms of traffic, just make sure you have two proxy servers on both sites. As to backup jobs and site A, then use backup server on site A to run backup jobs and perform item-level/file-level recoveries. This is one of the most widely used approaches among our existing customers.

If you decide to proceed with two backup servers, don't forget to install Backup Enterprise Manager and Veeam ONE (optional) to have visibility across all your servers via single GUI.

Thanks!

[wkuypers]

Thanks Vitaly.

But does this then mean I need to purchase additional licenses for the Backup Server at Site B to manage the replication jobs?

[veremin]

Hi, Willem.

It might be worth checking this sticky FAQ, since all of your licensing questions are covered in it.

Hope this helps.
Thanks.

[chrisdearden]

Thanks Vitaly.

But does this then mean I need to purchase additional licenses for the Backup Server at Site B to manage the replication jobs?

If you use enterprise manager to manage the licences, no remeber the licence is on the number of sockets you are backing up or replicating from . If you plan to back VM's up on site b then you would need to licence those hosts too.

There of course is plan B - replicate the backup server

[cdeal]

Hi Willem,

I would recommend to install a secondary Veeam B&R server on site B to manage replication jobs only. I don't believe there should be any difference in terms of traffic, just make sure you have two proxy servers on both sites. As to backup jobs and site A, then use backup server on site A to run backup jobs and perform item-level/file-level recoveries. This is one of the most widely used approaches among our existing customers.

If you decide to proceed with two backup servers, don't forget to install Backup Enterprise Manager and Veeam ONE (optional) to have visibility across all your servers via single GUI.

Thanks!

I'm new to Veeam (a future customer) and I have a similar setup with two sites (Veeam servers will be physical and proxies will be virtual).I thought I had read all the manuals and watched all the video's over the past several months trying to wrap my head around the best setup for Veeam to replicate offsite and possibly get Veeam backup's offsite as well, and this is the first time I have heard of this option. I think it will work well for my environment, but I want to make sure I understand the recommended setup:

1. Veeam server and proxy at the main site will handle backups to the onsite repository and to the DR repository using the off-site proxy at the DR site (offsite backups).
2. Veeam server and proxy at the DR site will handle replication from the main site (using the local proxy) to the DR site, as well as backups of VM's running at the DR site.
3. Enterprise Manager and Veeam One server at the main site will allow a consolidated view and reporting capabilities of both sites.

If this is correct, then I have a few questions: 1. What happens in the event of a fail-over to the DR site and the Enterprise Manager server is located at the main site, that's down. Will it be needed for any reason to get the DR site up and running 2. I'm trying to understand the fail-back process with this setup. Assuming the main site comes back online with the Veeam servers in the state they were in at fail-over, will the Enterprise Manager and Veeam servers, at the main site, be aware of the fail-over? Maybe a simple outline of the basic steps to fail-back, with this setup, would help me understand.

Thanks.

[Vitaliy S.]

Hi Craig,

Welcome to Veeam Community forums!

Yes, all of this is correct. If you place the secondary backup server on the remote site, you'll be able to failover to the replicated VMs via this server when the main site goes down. Of course, you can failover manually via vSphere Client, but in this case you will not have such features such as RE-IP and etc.

1. EM will not be needed to perform a failover. All you need is a backup server, which will be protected, since it is located on the DR site.
2. The Veeam server on the remote site will be managing the failover/failback process, so there is no need for EM and backup server on the main server to be aware about this. Keep in mind that EM collects information from backup servers regularly, so it will "know" about failed over jobs triggered by the second "live" backup server once the new collection job is triggered.

Hope this helps!