I can't discover FC LUNs from my storage on my host

[Zadok]

Hi

So straight to the point

I am unable to discover my LUNs on the server. I have the MULTIPATH IO feature installed (It is a Windows Server 2019)

I have a NetApp FAS2750 with a disk shelf and FCP license enabled.

I have created iGroups with the server WWN, and I have zoned the aliases created on the switch on the storage and servers. It is two Cisco MDS switches not configured in HA.

The two servers have FC ports connected on both switches.

Is there anything else I have to install on the Server to discover

-the LUNs provisioned

-the connectivity is correct?

I need help on this.

[TMACMD]

On the Netapp you can verify the host is logging in

 fcp initiator show

 you should see  the hosts and all the initiator wwpns show up if zoned correctly. 
if your igroups are correct the wwpns will be mapped the output 

you can also check the switch:

 slow flogi database

 you should see all initiators and targets logged in

 the biggest mistake I see: people use the hardware wwpns on the Netapp (that start with 50:) instead of the virtual wwpns that start with 20:

 on the Netapp: 

Network interface show

 you will see your Netapp Wwpns. Those should be the wwpns the zone on the switch

 make sure you also didn’t flip the fabric connections

[Zadok]

Thank you for your response

However, from this, I have no connected devices on the storage yet.

The fabric connections are fine, the servers are represented on each switch.

My issue is with the WWPNs to be mapped on the switch from the storage.

I have created an alias with the WWPNs of the storage, that start with 20: but I can't discover the LUNs still.

My concerns now:

- Is there a NetApp software I have to install to discover LUNs?
- If I can not discover the hosts from the storage, what is usually the most common issue?
- what's wrong with using the hardware WWPNs?

- those storage LIFs (starting with 20:), the WWPN is different from the WWNN. (20:0d, WWPN, became 20:00 WWNN). The WWPN is what actually appears on the storage. Is this correct too?

Thanks.

[TMACMD · Accepted solution]

For starters, you should read up at least a little bit on what you have. You will find out that every single part of a Netapp storage virtual machine (svm) is virtualized. Therefore you do not refer to any physical device rather virtualized/logical ones (like a LIF or logical interface)

do the “network interface show” command and look at the wwpns. Those are what MUST be zoned. If they are not showing up on your switch then you have not enabled NPIV. From the docs:

All of the switches that are connected to the fabric must support N_Port ID virtualization (NPIV) and must have NPIV enabled. ONTAP uses NPIV to present FC targets to a fabric.

if you don’t have that enabled then it will not work

 Netapp does have a SAN Host kit ( version 7.1 if I recall) that should be installed but won’t fix issue you have. All that does is add a few commands for troubleshooting a previously working environment and tunings to the host. 

use the correct wwpns (Not wwnn)

 on the switch, verify everything is logging in: show flogi database

 After you have enabled NPIV and zoned properly you do a: fcp initiator show

 you will see your hosts logged in even if LUNs are not available. The zoning allows the host to log in. If LUNs are available to the wwpns that are logged and mapped properly then the host will see them

[Ontapforrum]

I am hoping you have zoned "Logical LIF" FC WWPN, and not Physical WWPN.
https://kb.netapp.com/Advice_and_Troubleshooting/Data_Storage_Systems/FAS_Systems/Cannot_see_Fiber_Channel_(FC)_initiator_logged_into_the_cluster

[Zadok]

Yes. I did.

I removed the physical WWPNs.

[Zadok]

Okay

So I worked on this and found out that the issue was physical connectivity.

I fixed this, and then the fcp initiator show  command was able to discover all legs of the servers.

So I have also found the LUNs on the server.