Veeam, VPLEX and vMSC

[pageda]

From reading other posts, I can infer that Veeam works with VPLEX, but I'm looking for both confirmation and any reccommendations/best practices.

We're planning a vSphere 6, 3host per site, stretched cluster across two geographically separate datacenters with 10Gb interconnect, stretched layer 2 network. VPLEX will be in front of VNX arrays at both sites to provide active/active data. Total storage is currently small (about 4TB). I was planning on having a Veeam management server or proxy in each location?

I'm open to comments

[Gostev]

Veeam is completely storage agnostic, and will work with any storage supported by vSphere.

Given good connectivity in place, normally I would recommend a single Veeam backup server in HQ, and a backup proxy in each site. But, considering such a small size of infrastructure, you can as well use just a single proxy.

Thanks!

[dellock6]

I can confirm as I'm working these days with a customer that vplex is seen just like a FC storage, so you can run directsan backups against it with no issue.
I also suggest you deeply a second proxy at secondary site so when lun is served from that location the proxy doesn't have to cross the wan to grab data, even if the repository will still be likely at one location (unless some clustering is also planned for repositories)

[pageda]

Gentlemen, thank you both. I've been a veeam 'evangelist' for years and your quick, verified, expert opinions directly from company reps to consumers/users is invaluable and appreciated. Thanks to you both and the entire Veeam team.

[tinto1970]

good day, sorry to re open this topic after "some" weeks

I'm working on a customer infrastructure using VPLEX and experiencing this issue:

when the veeam proxy reads data in SAN mode with no storage integration, the speed is sometimes good (i.e. 50 MB/s per process, with many in parallel) but sometimes so slooooowww (I mean less than 500 KB/s).
Apparently it happens randomly, but this different speed is often observed for some vmdk inside the same VM (disk 1 is read fast, disk 2 pretty slow).

Any idea on what could be going wrong? Today I will try to get some logs and probabily open a SR.

Currently I have workarounded the issue forcing the proxy to work in NBD (management interfaces are on 10Gig and speed is satisfying for me: personally I would use NBD but the customer seems to prefer the direct SAN for traffic separation)

[foggy]

Hi Alessandro, opening a case is the right step with such issues - logs will tell exactly what happens during the slow processing of the affected disk.