The way we do it is an indirect approach. We have deployed the EMC EBR appliance (which is a VDP appliance mutation) and use it to back up our vRA VMs. vRA Users who want their server to be cacked up, can then use resource actions on their VMs, which in turn move the server to a resource pool that is backed up by EBR.
AFAIK a direct vRA - Networker integration is not possible. Also, the above mentioned solution is really rudimentrary and does not allow for automated, user-executed recoveries, since there is no API available for the EBR appliance.
At the moment we wait for Networker 9.2 to be deployed, because you can do a lot more with automation there.