Veeam WAN Accelerator vs Riverbed

[mzamesnik]

We have an environment of mixed Riverbed WAN optimization appliances, but all have plenty of cache space. My question is, what are the effects of having both WAN Acceleration, and optimization through the Riverbed for Backup Copy jobs over a very slow link (512K - 1.5Mbps)?

Could the Riverbed be inhibiting performance in any way by performing its optimization in conjuntion with the built-in WAN acceleration of Veeam? Is there any documented best practices for setting up WAN acceleration in Veeam, and *why* it seems to be recommended to bypass existing optimization appliances? I am interested in this from the standpoint of the effects on the Backup Copy job performance, not the performance of the Riverbed using CPU cycles and cache space, which I am being told is not a concern.

[Gostev]

Bypass is recommended solely because Riverbed cache will fill up with our backup data, and this will slow down the user and production app experience because of no cache hits, thus not letting you achieve the main reason why people deploy general purpose WAN accelerators.

I don't expect Riverbed impacting the Backup Copy job performance. I also don't expect it to deliver any significant benefits. However, adding more data processing into the picture does increase risks of data corruption accordingly.