Delete and Recreate a Service Profile using the same address (HBA WWNP) from WWNP Pool

[lomonaco]

Hi guys,

   I have 8 services profiles created used by esxi boot via SAN. Cause is boot via SAN is very important to me the HBA WWNP Address.

   Now I need to modify the vNIC Template used in those services profiles to include the VM option in the Target.

   But to do that I will need to delete and recreate the vNIC Template.

    I think the UCS will not permit me to that until I disassociate and delete the 8 services profiles ( I can be wrong)

    My question is.

     After I delete a service profile, whats the better way to garantee that the new new service profile will be used the same address (HBA)

     used by the service profile deleted. ?

     Thanks in Advanced,

      My Best Regards,

[Saurabh Kothari · Accepted solution]

Hey,

I am not sure why you will have to delete the 8 SPs when the changes(deletion) are to be done on vnic/HBA template.

WWNP address is being derived from a pool? right?

what is the assignment order? Default or sequential 

is this WWPN pool specific to only these 8 Servers?

I did a quick test when I delete HBA template it didn't delete the vHBAs configured in SP.

thanks,

-Saurabh

[lomonaco]

Hi guys, thanks for the reply.

The WWNP is being derived from a WWPN default pool (unfortunately no sequencial)

I'm think about to delete the services profiles (only 8) and recreate the pool as sequencial

Thanks again for the answers

[Walter Dey]

Perfect ! and please don't use any default pools ! create specific ones, and make them all sequential.

[Walter Dey · Accepted solution]

There is no other choice: you have to delete the service profiles; vhba's take pwwn / nwwn from the defined pools (I hope you made them sequential); the pwwn... reservations stay, as long as the SP exists (SP could be associated or disassociated, doesn't matter).

Therefore delete the existing SP's; make the change ! create new SP's; and as long as the number of vbha's stays the same, your pwwn/nwwn are the same as in the beginning.

Of course this is disruptive; it could be done server by server (or SP by SP) which is acceptable in a VMware environment.

Walter.