Backup and Replication jobs overlapping

[Dave23]

I have 2 sites; a production site and a DR site. I am fairly new to Veeam B&R and had set up a B&R server at the production site with a local repository for backup jobs and a separate B&R server at the DR site for replication jobs. I only have one vCenter server which is located at the production site which manages all the ESX servers including the one at the DR site. I have multiple proxies at the production site and one proxy at the DR site. I run continuous replication on my critical VMs and perform daily backups after hours. I've been running into issues lately that I think are related to the backup and replication jobs overlapping each other since they run on separate B&R servers and are not aware of each other.

Would it be best practice to just have one B&R server at the DR site manage all backup and replication jobs? I would still only use proxies at the production site and a windows server at the production site for the repository. Then use that same B&R server at the DR site to manage the replication jobs. This way, the B&R server is aware of all the jobs and they will not overlap. Is this what others do in the same situation? It was suggested by one support engineer to set it up the way I currently have it, but the jobs have many errors and actually crashed one of my VMs because both servers were managing snapshots and locked one up. Is there a downside to managing your backup jobs from a different site if the proxies and repository will be at the production site with the live VMs?

Thanks,

Dave

[Vitaliy S.]

Hello Dave,

Yes, managing your entire infrastructure from the DR site will "resolve" your current issues with jobs overlapping, however can you please tell me what exactly does not work in the current scenario?

Thank you!

[veremin]

Additionally, you can utilize PowerShell in order to check the status of the jobs that are executed at Production site before the DR jobs will be started. The script will connect to VB&R instance in DR site, using invoke commands. Then, check the status of the corresponding job using .GetLastState()method and execute the required backup jobs, only if the status isn’t equal to “Working”.

Thanks.

[mrstorey]

I have configured all of our sites to use backup servers / job schedulers at their corresponding failover site for this very reason.

So, site 1's backup server manages site 2's backup, replication and backup copy jobs using proxies at site 2, and vice versa.

Seems to work well, and was the only way I could figure out to achieve a low RPO with replication jobs while ensuring backups were successful and always completed.

The only downside I can see so far is that Exchange and Sharepoint explorers must be run local to the repositories, because unlike normal restore jobs they can't seem to use a remote proxy (happy to be corrected otherwise!) - just means depending on the type of restore you're doing, you'll be performing it from different servers.

[veremin]

Just noticed that you’re talking here about continuous replication. In this case, the only option you have is, indeed, to manage everything (backup/replication) from the DR site. This way, the backup server will be aware about overlapping jobs and queue them appropriately. Thanks.

[kbhosale]

You can schedule a job for particular time duration using Backup window option. I am also using similar kind of Veeam architecture for backup and replication and I am using backup window option and select the time duration for backup job and replication job on both BnR servers.

[getzjd]

Additionally, you can utilize PowerShell in order to check the status of the jobs that are executed at Production site before the DR jobs will be started. The script will connect to VB&R instance in DR site, using invoke commands. Then, check the status of the corresponding job using .GetLastState()method and execute the required backup jobs, only if the status isn’t equal to “Working”.

Thanks.

Hello, I was wondering if you could post an example of the script to run before a replication to check a remote site to see if the backup has completed? Forgive my lack of powershell skills

Thanks

[veremin]

You should use invoke expressions to connect to production site. Then, it's just matter of checking the state of replication job:

Code: Select all

$Job = Get-VBRJob -name "Name of replication Job"
$Job.GetLastState() -eq "Working"

Thanks.

[getzjd]

Thank you. Unless someone else can assist with a little more direction on this, I will try to piece something together although my knowledge of PS scripting for Veeam is minimal. Essentially, I need a script to run before the job, that checks a remote site to ensure a replication job has completed. This will be helpful where a replication job may run for an extended period of time due to a larger incremental transfer caused by Windows 2012 deduplication of a file server.

Basically, Backup job at Site A starts, executes a script that will check the remote site B to ensure the replication job is not running. It will then keep the backup job at Site A from executing until the replication job is complete.

I understand this piece, but I need some sort of if, then loop to execute that continues to check the status of the remote job say every 60 minutes and once it sees the status as not working, then will allow the current job to proceed.
$Job = Get-VBRJob -name "Name of replication Job"
$Job.GetLastState() -eq "Working"

[veremin]

Please, create a separate topic in the corresponding PS subforum in order not to overfill the existing discussion. Thanks.