Replication Speed Drop

[huntermaz]

We have been using Veeam for years, about two weeks ago our replication jobs when from around an hour to 6+ hours. We have testing out connection between the two sites and we are getting the same speed as before(50mb). It is now showing the target as the bottleneck now. How do we troubleshoot this? Besides some Windows and VMware patches nothing has changed. We are using a Virtual Appliance on each end and we have Equallogic SAN's on both sides. I think this job has been inplace for about a year. Last time we had a slow down it was the WAN connection but we did check that.

[foggy]

Nathaniel, please check the transport mode being used on the target side (do you see the [hotadd] tag next to the target proxy server name in the VM processing log?) and target datastore latency.

[Vitaliy S.]

Also please look through the session log. What operation takes most of the time, data transfer or something else?

[huntermaz]

Getting hot add is not supported for target disk, failing over to network mode. Where do I find the session log?

[foggy]

Getting hot add is not supported for target disk, failing over to network mode.

If previously hotadd was used to populate target datastore, then this is the reason of slow down. Check whether proxy server still has access to the datastore. This KB article can also help in identifying the reason of hotadd failure.

[huntermaz]

So it ended up being that the proxy didn't release the Hot Add disks. There was 15 disks attached to the proxy. I removed all the ones that shouldn't be there and ran the replication job and it took an hour. So back to normal.

[foggy]

Looks similar to this issue. Glad you're sorted now, however if the behavior shows up again, don't hesitate to contact technical support for investigation.