Originally posted by: anarasim
I am looking for best practices for Host Path failover when doing switch maintenance updates (Brocade 4100) or storage microcode updates (NSC/AMS).
We have redudant paths for ports in NSC55 and the HBAs in hosts (Solaris Servers running Veritas DMP path failover software) connected to 2 x Brocade 4100 switches in the SAME fabric. If we have to do switch/storage maintenance, do we have to to do a manual failover on one path, do the maintenance on the components in secondary path and then do a manual failback to do the maintenance on the components in primary path or can the maintenance activites be carried out online assuming an automatic path failover/failback at the host end? What is the recommended best practice?
Does the host path failover functionality differ for USP/NSC (active-active) Vs AMS (active-passive)?
Does the path failover using VxDMP differ from that of HDLM?
Your inputs are greatly appreciated.