why do we have to manually add WWPNs for FC hosts on system manager?

[FelixZhou]

configured a few ESX hosts, have to manually add WWPNs on system manager, eventually rebooted esx host to see devices. Why we have to manually add WWPNs? is that caused by host compitability or there is any way work around? thanks

[SpindleNinja]

They're used in conjunction with igroups so the storage / (ONTAP) knows what disk(s) to hand out (allow access) to what host(s)

additional reading for you. 

https://docs.netapp.com/ontap-9/index.jsp?topic=%2Fcom.netapp.doc.dot-cm-sanag%2FGUID-6B12E8A0-A101-4106-AFD6-77826CFBDC25.html

https://www.netapp.com/us/media/tr-4080.pdf

[TMAC_CTG · Accepted solution]

If I recall for both FC and iSCSI:

If you setup Zoning on the FC switches (for FC only) and/or point the iSCSI hosts at the NetApp for login

then I believe things like "fcp initiator show" and "iscsi initiator show" whill show all FC and/or iSCSI initiators talking to ONTAP.

When you go to the GUI/System Manager and try to create an initiator group, I believe it does in fact pull up any WWPN or IQNs that are talking to ONTAP. Now, with WWPNs, what you DO NOT have is which host belongs to to which WWPN. iSCSI typically has the hostname in the IQN so it is easy. With FC< you really need to know what hosts the WWPNs belong to.

[FelixZhou]

in my case, after we made zones, went to GUI for creating igroup, we didn't see host initators in pull down list, we manually added the wwpns for hosts,  we ran "igroup show -igroup <xxxx> -instance", all ports showed "not logged in", eventually we had to reboot our hosts to see devices on hosts. Not sure if we missed any steps.

[TMAC_CTG]

Two things here

 1. after the zones are created on the switch, you should be able to do something like "show flogi database" on a Cisco MDS and see which targets and initiators are logged in. After verifying the zones on the switches if the GUI is open you may need to force a refresh to have the GUI see the WWPNs in the pull down. They do show. But a GUI refresh may be needed

 2. on the hosts the Trevor was not needed either. After the zones are set and the igroups are created and the LUNs are mapped the host only needs to rescan the FC bus for new targets. When you rebooted it forced a scan anyway

[SpindleNinja]

i've seen it behave like that before.    did you rescan after your added the initatiors in vmware?     

[FelixZhou]

i am pretty sure we refresh the GUI, a few people checked on different system manager consoles.