Assistance with DD/VBR v8 Architecture

[brupnick]

Good morning-

I wondering if someone could provide some assistance with my current architecture and what I'm trying to achieve.

At my main site, I have a VBRv8 installation and a DD7200. At my DR site, I have a separate VBRv8 installation with a separate DD. Jobs are scheduled to run at my main site and write to the main DD, then DD replication sends all of my backup files from main to DR.

In the event of a disaster, my intention was to import the backups that were replicated from the main site DD to the DR DD into the DR site VBRv8 installation. Unfortunately, I'm having trouble finding out how to do this. If I add the DD to VBR via CIFS, this repository is not in the list of sources available for import. If I mount the DD to my VBR server via CIFS or NFS, these drives are also not available for import. Is my only option to spin up a Linux VM and mount the DD via NFS and add the Linux server to my DR VBRv8 installation?

I really hope that I'm just missing something silly here and will have a "D'oh!" moment when someone points it out to me.

[tomb80]

Its going to depend on how you set up the repository. We are using Veeam 8 with a DD2500 and a DD860.

On Veeam 7, we were using just the 10GB network since no DDBoost. What I did was set up a replication set so it was replicated over. After the replication completed, I added the DD at our DR site as a backup repository to the Veeam 7 server there. After that, you should be able to import backups. I noticed that once you rescan a repository in Veeam 8 that it automatically adds your backups in to Backups\Disk (Imported).

I had OK backup times with Veeam 7 and I expected a lot better backup times using DDBoost over 10GB or DDBoost over FC, but our backups are painfully slow with the DD platform. I am not sure whether it is Veeam, the settings of the jobs, or the DD appliance itself. I have a ticket open for this issue. 00677553.

Are you using DDBoost with your DD7200?

[brupnick]

Thanks, Tom.

Yes, I am using DD Boost with my DDs and VBRv8.

What I did was set up a replication set so it was replicated over. After the replication completed, I added the DD at our DR site as a backup repository to the Veeam 7 server there. After that, you should be able to import backups.

How did you add them in v7? CIFS or NFS?

For me, my backup times are pretty good and my bottleneck is never the target (DD). I'm waiting on cabling so my physical proxies can be hooked directly to my SAN which should help balance out the bottleneck which is consistently my source.

Thanks!

[foggy]

If I add the DD to VBR via CIFS, this repository is not in the list of sources available for import.

If you open the Import dialog, you can just browse to any network location and select backups to be imported (just leave "This server" selected in the drop-down list and click Browse). Alternatively, when you add a folder as backup repository, backups are imported automatically during repository rescan procedure.

[brupnick]

And it really is that simple. Thank you so much, foggy!!

[smpartridge]

I have a similar question regarding the design of DD/VBR. We have a DD2500 at both the HQ and DR site. There is a 250MB connection between the sites. HQ has a NetApp8020 and the DR is a NetApp3020. There are 11 ESX5.5 host at the HQ with 2 ESX5.5 at the DR. All ESX host are the same. I have the ability to run physical, virtual, or a combination of both for VBR. My question is what would be the best number and configuration of servers for Veeam? Any insight or experience is helpful as we are looking to migrate from vRanger.

[veremin]

Hi, Shawn,

Are you going to run replication jobs to the second site? If so, the recommended approach is to deploy additional backup server at that site and make it orchestrate replication activity. If something happens with the first site, you will be able to execute DR procedures from second site smoothly.

Also, don't forget about protecting VB&R server, using configuration backup.

Thanks.

[tomb80]

I am doing CIFS and NFS in V7. We started with CIFS primary, but moved over to NFS. We saw a slight bump in performance there but not much. DDBoost has been questionable at best. There is no reliability for it. I have jobs running on CIFS/NFS on v7 that take 6 hours. With DDBoost over network or FC, they take up to 31 hours.

[SGalbincea]

I am working on figuring out how to replicate the backups from our HQ DD2500 to our DR DD2500. We currently have a 1Gb link.

I like the efficiency of the DD replication, but cannot for the life of me figure out how to replicate a DDBoost SU from HQ to DR and have both sides accessible via DDBoost. If I set up HQ as a DDBoost SU and then tell it to replicate to DR, it creates a new MTREE at DR which cannot have DDBoost enabled (seemingly). If I create the DDBoost SU at DR first and then attempt to set up replication it fails since the MTREE already exists...ideas?

Additionally, I am not so sure I want the DD doing the replication as if there is corruption in the backups at HQ it will immediately be replicated to DR, so I am exploring the possibilities of using Veeam to handle it instead...but so far performance has not been very good.

Any suggestions or insight into how you guys are doing this would be great.

[dwclarknu]

Were you able to find a solution to this? We are in the same boat.

Good morning-

I wondering if someone could provide some assistance with my current architecture and what I'm trying to achieve.

At my main site, I have a VBRv8 installation and a DD7200. At my DR site, I have a separate VBRv8 installation with a separate DD. Jobs are scheduled to run at my main site and write to the main DD, then DD replication sends all of my backup files from main to DR.

In the event of a disaster, my intention was to import the backups that were replicated from the main site DD to the DR DD into the DR site VBRv8 installation. Unfortunately, I'm having trouble finding out how to do this. If I add the DD to VBR via CIFS, this repository is not in the list of sources available for import. If I mount the DD to my VBR server via CIFS or NFS, these drives are also not available for import. Is my only option to spin up a Linux VM and mount the DD via NFS and add the Linux server to my DR VBRv8 installation?

I really hope that I'm just missing something silly here and will have a "D'oh!" moment when someone points it out to me.

[foggy]

Have you tried to perform what I've described above?

If you open the Import dialog, you can just browse to any network location and select backups to be imported (just leave "This server" selected in the drop-down list and click Browse). Alternatively, when you add a folder as backup repository, backups are imported automatically during repository rescan procedure.

[rgarrison]

We have been able to successfully replicate our Veeam DDBoost backups to our DR site with Data Domain built-in replication. We tried using Veeam backup copy jobs, but it was just too slow over the WAN.

The key is setting up mtree replication of the SU to the DR site, then attaching the DDBoost user account to the SU on the remote (DR) Data Domain. Once that is done, you can add the DR DDBoost SU as a DDBoost repository in VBR. Perform a rescan on the remote repository, and the backups will show up in the imported section.

This will NOT address your concerns over a corrupted backup being replicated, but it will get your data moved to the DR site in a much more efficient manner than Backup Copy is currently capable of doing.

ETA: EMC support told us we needed to be on DD OS 5.5.x for this to work. Not sure if 5.4.x would work.

[SGalbincea]

I think I have solved the corruption issue, at least for us. The DD has the ability to snapshot the MTREE that is being replicated, and in turn replicate that snapshot to the other DD. I have confirmed that I can mess up the .vbk on the primary copy and the replica, yet still access the backup successfully by mounting the snapshot on the DR side. As of last night I have disabled our copy jobs and am now using DD replication to copy our backups to DR.