ESX replication job optimization

[rxebk]

Hi
New to the replication tool, I would like to know if there are a few things to consider when setting up the replication job:
At the moment, I have three jobs running, one is running at 24MB/s, one at 5MB/s and one 3MB/s.

We have a 100MB link between two physical sites.

The difference between the three jobs: they have a different number of VMs in the job. Do I need to consider having less VMs per jobs maybe?

Does the number of CPUs impact the replication process like the backup does?

Any advice would be good.

Thanks

[Vitaliy S.]

Hello Eric,

First of all, I would make sure you've specified service console credentials for your target ESX host in the host connection settings, that would give you a better performance rates for replication jobs. Secondly, could you please clarify what replication mode you have chosen?

The difference between the three jobs: they have a different number of VMs in the job. Do I need to consider having less VMs per jobs maybe?

How many VMs do you have now added to your replication job? You may split those, as replication doesn't have the same benefits of compression and deduplication compared to backup jobs.

Does the number of CPUs impact the replication process like the backup does?

Yes, please make sure CPU usage is not maxed out during your replication jobs.

Hope it helps!

[pmalinov]

Hi
New to the replication tool, I would like to know if there are a few things to consider when setting up the replication job:
At the moment, I have three jobs running, one is running at 24MB/s, one at 5MB/s and one 3MB/s.

We have a 100MB link between two physical sites.

The difference between the three jobs: they have a different number of VMs in the job. Do I need to consider having less VMs per jobs maybe?

Does the number of CPUs impact the replication process like the backup does?

Any advice would be good.

Thanks

Also keep in mind that 100 Mbps is not 100MB/s.

Telecommunications bit rates

Bps = 1 bit/s
Kbps = 1.000 bits/s
Mbps = 1.000 Kbits/s or 1.000.000 bits/s
Gbps = 1.000 Mbit/s or 1.000.000.000 bits/s
Tbps = 1.000 Gbit/s or 1.000.000.000.000 bits/s

Data Storage

Byte = 8 bits
Kilobyte = 1024 bytes or 8192 (8 x 1024) bits
Megabyte = 1024 Kilobytes or 1.048.576 (1024 x 1024) bytes or 8.388.608 bits
Gigabyte = 1024 Megabytes or 1.073.741.824 bytes or 8.589.934.592 bits
Terabyte = 1024 Gigabytes or 1.099.511.627.776 bytes or 8.796.093.022.208 bits

[rxebk]

Thanks for your answers. Do you know if the application is multithreaded? I have 2 CPUs dual cores and they are heavily used which I think is where the bottleneck comes from at the moment.

What would be the recommanded configuration to replicate 15 servers composed of 13 servers between 20GB-100GB and two servers 300GB each?

I am thinking on splitting the backup jobs on one server and get another server dedicated for the replication with 2 CPUs 4 cores.

Many thanks
Eric

[Vitaliy S.]

Eric,

Yes, we are using all available CPU resources. Generally, I would recommend setting up a 4 vCPU VM with Veeam Backup server installed and use VA mode to replicate your VMs. For more info on this mode, please review our sticky F.A.Q. section.

Thanks!