V6 and legacy replication job performance

[mmartin]

Hi

Probably something simple but can't figure it out.

We have a single physical veeam server running V6 - all replication jobs are legacy jobs.

We are replicating to another ESX on the same Lan but we are trying to size the link using Hyper IP -
Proxy is on the same server as Veeam and no proxy on the DR ESX.

Veeam box - 10.0.0.5
Hyper IP - 10.0.0.3
Hyper IP DR - 10.0.0.4
DR Host - 10.0.0.242
Static routes are setup on both Veeam box and DR host and communication is working fine.
1000 base Network connections
Ok so the issue.

After I upgraded and when I run a replication job the job is taking forever - I have the hyper IP box restricted to 6mb however the hyperIP box is only recieving max 3mbits / s - can be seen in the graphs and the stats.

The hyper IP stats should show the pipe being filled to 6mbits but its not, I spoke to NetEx but they said the hyperIP are operating fine just not receiving the data from veeam.

So this leads me back to Veeam.

Any ideas.

[Gostev]

Hi Michael, I would advice that you re-create your replication job to leverage new replication engine (otherwise, what was the point of doing v6 upgrade in the first place). As you can imagine, legacy replication engine was not the focus of this release, so they may not produce optimal performance along with other engine changes, and it did not have enough QC focus to spot this (just guessing). However, if you really want to spend time troubleshooting the discontinued replication architecture, then please follow the regular procedure and open the support case - as obviously it is impossible to troubleshoot anything from the short text above (full logs are needed, as usual). Thanks.

[mmartin]

Hi

I had fully planned on changing the jobs however I did not expect the performance to drop after the upgrade.

Thanks though I will log a support ticket.

Michael

[mmartin]

another question in relation to this - I noticed that my NIC is fully utilised sitting at 95 - 98%

We are using Network mode for backup & replication jobs as this is a seperate veeam server.

One of our backup jobs has just started also and I noticed slow performance with it and this is backing up to the local disk in the veeam server so I am assuming the bottleneck is the nic...

Any thoughts on how to get around this, I don't have direct SAN access nor can I use VM mode as this is a physical box.

[Gostev]

May be some NIC or switch on your network just failed over to 100Mbit because of some physical/cabling issue?