Replication Of Various VMs

[ituhp]

Hi,
We have 2 physical hosts and are replicating from host01 to host02.
We currently have 1 job which does several VMs including AD Controller, SQL and a server which mostly runs Services.
We have a single job set to continuous replication so that we can failback to latest possible in case of emergency.
The AD and Services server`s replication finishes much quicker than the SQL.

I am curious as to whether we would be better separating the faster replicating servers and the slow (SQL) server and running 2 jobs...
I am intending on adding a file server into the mix, and assume that this one will be a slow replication also due to volume of data.

Any advice would be greatly appreciated.

[Mildur]

Hello

You can define a processing order in the same job, if you add the VMs as standalone VMs. Put the File and SQL Server at the end of the job. That will make sure, that the other two servers are replicated first.
https://helpcenter.veeam.com/docs/backu ... ml?ver=110

If you expect the fileserver to have a long initial replica duration, then it's probably better to use a separate job or use replica seeding if you have already backup files at the second location (if you use one).
https://helpcenter.veeam.com/docs/backu ... ml?ver=110

A second Job will not make things faster, if you have limited bandwidth between both Hyper-V Hosts and to much incremental data to transfer. Probably a WAN accelerator could help in that case.

Thanks
Fabian

[ituhp]

Thank you for that.
I understand that can be done, but was trying to see if different jobs would be better.
If I have 2 small and 2 big servers in a job, then the 2 small ones need to wait for the big ones to finish before next replication, right?
would having 2 jobs deteriorate the performance of the replication?

[Mildur]

If I have 2 small and 2 big servers in a job, then the 2 small ones need to wait for the big ones to finish before next replication, right?

Yes, they will. In that case a second Job would allow the 2 smaller ones to run, if there are enough tasks on the proxy server available.

would having 2 jobs deteriorate the performance of the replication?

Both Jobs will share the same component (hypervisor, production storage, network connection, Proxy Server CPU and memory). If for example the network bandwidth or read performance from the production storage is already to low, then 2 jobs wouldn't make it faster.

What bottleneck do you see in the job statistics?