Traffic Throttling not Applying to Backup Copy

[jacktwomey]

Hi There,

I've read through the forum and although there are similar posts, none of the responses have resolved our issue so I hope people don't mind me posting again.

We have 2 x Veeam B&R servers either side of a site - site VPN connection. We're running a backup copy via WAN accelerators in both directions over the VPN connection. Site A to site B the traffic throttling works fine. However site B to A the throttling doesn't want to know. Both sites have different LAN subnets, but appropriately are set within the Source and Target fields.

Both Veeam servers are identical (V7 patch 3). I've gone as far as to totally re-install the Veeam server at site B - still it doesn't want to know.

We're trying to throttle the copy to run at 6Mbps but without the throttle it'll spike up to 100Mbps (which we'll get charged for with our bandwidth commit rate at the DC). Currently we've applied a throttle on our firewalls, but this workaround is nowhere near as ideal at the throttle function within Veeam.

Any ideas welcome!

Thanks

[foggy]

Jack, first of all, please tell us how do you configure traffic throttling rules. Basically, throttling rules are configured within a single Veeam B&R instance and are applied between two Veeam data movers running on specific IP's (be it either proxy, repository server, or WAN accelerator) and they are unaware of the components that belong to the second Veeam B&R instance. Also, throttling rules are reversible and each rule should function in both directions.

Do you configure backup copy jobs in different directions on different Veeam B&R instances? Do you configure corresponding throttling rules on both Veeam B&R instances?

[jacktwomey]

Amazingly, after battling this for about 6 hours yesterday I seem to have resolved this 30 minutes after posting on here!

I entered another rule which is a mirror to what should have been required, re-started the copy job and hey presto - it's now throttling the job perfectly, with the Bottleneck even displaying as "Throttling".

What is interesting is firstly why entering the rule the wrong way round, has made this work. Secondly your point "throttling rules are reversible and each rule should function in both directions" certainly doesn't appear to be the case.

My setup is basically site A on a 192.168.8.0 network and site B on a 172.30.8.0 network. The rule at site B (which was not working) was source: 172.30.8.1 172.30.8.255 and target: 192.168.8.1 - 192.168.8.255 . Once I added a second rule; source: 192.168.8.1 - 192.168.8.255 and target: 172.30.8.1 - 172.30.8.255 , all started working.

To answer your questions:

"Do you configure backup copy jobs in different directions on different Veeam B&R instances?"
Yes so each Veeam B&R instance only has 1 backup job going in 1 direction.

"Do you configure corresponding throttling rules on both Veeam B&R instances?"
Yes - site A only required on rule. Site B required 2 (and as mentioned oddly in the wrong direction to work).

As mentioned this is working now, but would be interesting to know if we're doing something wrong!

Thanks

[foggy]

What is interesting is firstly why entering the rule the wrong way round, has made this work. Secondly your point "throttling rules are reversible and each rule should function in both directions" certainly doesn't appear to be the case.

The behavior you're describing looks unexpected, so if you're willing to dig down to the reason, I recommend to contact technical support so they could take a closer look at your setup and review log files, if needed.

[jacktwomey]

I may have spoken too soon... It's definitely better but not solved. The throttling keeps traffic at it's set speed, then every 1-2 minutes there are then huge bursts lasting up to 30 seconds which go way beyond our set throttle limit.

Foggy - going back to your first reply, can i run by you please... In our instance the throttle rules will be going via the WAN Accelerators. B&R server at site A is using it's local WAN accelerator and the remote which is on site B's B&R server for it's copy job. Then the server at site B is using it's local WAN accelerator and the remote which is on site A's B&R server for it's copy job.

I'm just wondering the fact that the B&R servers are sharing their local WAN accelerator both for their local copy job, and as the partner of the alternate B&R server? Could this cause an issue do you think?

Thanks
Jack

[foggy]

Generally speaking, it should not. That's why I encourage you to contact support directly.

[jacktwomey]

Right you are, thanks for your time.

[foggy]

I'd be also interested in your findings regarding this, so please keep us posted.

[jacktwomey]

Shall do.