wan accelerator on veeam 9

[garibaldi]

Hi,

Anyone have experience using veeam wan accelerator for replication for big vmdk?
We have this VM with 1 TB disk that takes forever to complete the replication with wan accelerator enabled.

When we do not use the wan accelerator, the process completed with no issues.

Already checking the disk for cache, there are still space for cache in the source and the target.

Anyone can advice, any similar issues that happen ?

[foggy]

What kind of link is involved in your replication process?

[garibaldi]

Hi Foggy,

We are using ipsec (vpn over internet) to connect both sites.
Other VM that have size < 500 GB works perfectly with the wan accelerator.

[foggy]

I meant link speed mostly, sorry. Does the job for other VM use the same pair of WAN accelerators?

[garibaldi]

Mostly around 10 Mbps and it is shared, not dedicated.

Usually replication start at off peak hours (non office hours) and only one replication per day (RPO = 24 hours)

[foggy]

So am I understanding right, job incremental run involving WAN accelerators for this VM cannot complete (btw, what last operation do you see in the job log?) while direct run completes fine (though you'd like to lower traffic to have faster jobs)? What are the bottleneck stats for both jobs?

[garibaldi]

Process stuck at 3% and after 48 hours waiting, we decided to cancel the job for this VM.
Bottleneck status was source 20%, source wan 29%, network 24%, target wan 34%, target 11%.

[foggy]

I'd let our engineers to take a look at what job is actually doing during this time.