HPE EVA Storage

Best-practice zoning P2000 for failover

Occasional Contributor

Best-practice zoning P2000 for failover

I am about to configure a P2000 (dual controller FC/iSCSI version) but am aware that HP have changed tactics re: FC cabling in comparison to earlier MSA2000's so want to ensure I zone things correctly.


There will be two SAN fabrics (HP/Brocade 8/8 base switches) and I know that port 0 on each controller should got to one SAN switch and Port 1 to the other (different to earlier versions which suggested the ports should be crossed-over).


However, I want to ensure that I create correct FC zones.


So, assuming a single initiator/single target config, should I zone each server port to each of the four WWPNs of the P2000, on each of the switches? In other words, is there ever a situation where a "virtual" instance of a WWPN can move from port 0 to port 1 of the P2000?


I've spent a lot of time searching for this specific scenario but haven't yet found a definitive answer so though I'd try asking here.





Steven Clementi
Honored Contributor

Re: Best-practice zoning P2000 for failover

I don't even think that is possible... which is why it is called a WWPN (Port Name) vs the WWNN (Node Name).  WWPN's are unique across all devices. (or should be).


I was told today that the "Best Practice" has chnaged to now be Single Initiator/Multiple Target... and I would assume that "Multiple Target" means WWPN's from the same Array.  I.E.  You create a Zone with the host, controller A Port 0 and Controller B Port 0 in it.


Not sure how true it is... but I've been doing it for years ;o)


Steven Clementi
HP Master ASE, Storage, Servers, and Clustering
MCSE (NT 4.0, W2K, W2K3)
VCP (ESX2, Vi3, vSphere4, vSphere5, vSphere 6.x)
NPP3 (Nutanix Platform Professional)