Copy Job Proxy/Gateway allocation with CIFS Repositories

[pirx]

Hi,

I'm wondering if I have the right configuration for our copy jobs.

Environment:
- stretched VMware cluster over 2 sites, 40 Km distance, 0,5ms latency
- 2 physical Veeam proxies with 20 cores at each site
- at each site one NetApp filer (CIFS Shares) with one SOBR with 100 TB volume/extend for primary backup, one SOBR with 5 x 100 TB extends for copy jobs
- backup jobs to site A and site B
- copy jobs that transfer backups from site A to site B

Should I...

- set a static gw servers for each SOBR extend (I think I must because in my first try it used the B&B VM as gateway for all jobs and no physical proxy...)?
- use a proxy that is at the same location as the repository?
- set proxy affinity so that a extend only uses the 2 proxies that are local?

Because of the latency between the repositories I think it might make a huge difference how the data flow is and I currently don't know how to set this properly. On the other site, if I limit the number of proxies that can be used I restrict the max. number of possible tasks.

[pirx]

the connection between the sites is 20 Gbit/s
...
- change the number of multiple upload streams per job in the proxy network traffic rules section (default 10)