veeam scale out repository rebalance slow

[mathschut]

Hello, as soon as we start a rebalance, we see that all traffic for the rebalance goes through the Veeam Managent server. Since this is on a different network than the repo server, this process is very slow. Can you somehow change the whole thing so that the rebalance traffic doesn't go through the management server?

[Mildur]

Hello Mathias

May I ask, what sort of backup storage do you use as the performance extends?
Traffic should flow from data mover on one extend to the data mover on the target extend.

Best,
Fabian

[mathschut]

Hi, I have 4 Linux repos. According to Veeam, traffic always goes through the Veeam Management Server. But it is behind a firewall with only 1G connection.

[tyler.jurgens]

Are your Veeam repos on the same subnet? Can the repos resolve the DNS for each other? Can they ping each other? Do you have the required ports open? What did Veeam support have to say?

[mathschut]

Hi, all Veeam repos are located next to each other on the same network and can be reached via DNS and ping. Veeam support says that the traffic during balancing always goes through the management server.

[Mildur]

Hi Mathias

Veeam support says that the traffic during balancing always goes through the management server.

Can you please let me know the case number? I want to follow up your question with our QA team, but I require the case number first.

Best,
Fabian

[mathschut]

Yes of course.

Case-Nr.07135890

[Mildur]

Hi Mathias

My assumption from my first comment in this topic is confirmed by our QA team.
Rebalancing is directly done between your Linux repository extends. The management server is purely involved for management traffic. Meaning, the backup server has to talk to each extend to orchestrate the rebalancing. But backup data then will then be copied from one Linux extend to another.

There is one exception from this policy. If you have a repository which requires a gateway server (Data Domain, StoreOnce, NFS, SMB or object storage) and you have specified the backup server as the gateway server (or automatic gateway), then we may use the backup server as a gateway server to rebalance the data.

If you see something else, please open a new support case to investigate the behavior. Unfortunately your old case is already closed and no logs were provided. A new case is required. If you open a new case, provide us with a log package in the support case --> https://www.veeam.com/kb1832

Best,
Fabian