make simplified remote restart fully disaster recovery capable

When employing simplified remote restart, one requisite to restart an LPAR on another machine is that all SAN paths and targets (e.g. LUNs) have to be available on target VIO servers. During a disaster recovery scenario, this is not the case; in a fully redundant setting with two storage systems, two (or more) SAN switches and two power machines evenly distributed accross two sites, simplified remote restart does NOT work for a site failure. In this scenario, during site failure, multiple things fail at once: one power system, half of the SAN switches and one storage box (thus: half of the SAN paths and half of the LUNs) fail, and all LUNs of one storage will fail. However, if you have a LVM-mirrored AIX box, it could be able to boot with half the paths and with half the LUNs. But unfortunately, simplified remote restart fails because the number of SAN paths and SAN targets is not the same on the target machine. 

There is no way to override this behaviour, it is mandatory to have the same amount of paths and targes on source and target VIO servers / machines. 

To make simplified remote restart fully DR capable, there should be a way to override this behaviour and make simplified remote restart ignore path errors. It is better to have a chance to boot with failed paths and disks than to have no chance to boot an LPAR at all.