LUN access failure + blade not booting

Mich El · ‎11-08-2017

Hello All,

I am booting UCS-B blades from the external storage but 1 blade out of 8 fails. I got it logged to the fabric (as well as the target). Also I have default-zone permit for these VSANs on both fabrics A/B. When I check on the adapter level, I'm getting access failure for LUN0. But all LUNs are listed from the target to the initiator! (that's why I think this is not zoning related issue). Rest of blades are booting with no issues.

The output from one side (fabricInterconnect-a):

adapter 1/1/1 (fls):21# lunlist
vnic : 15 lifid: 4
  - FLOGI State : flogi est (fc_id 0x1e0141)
  - PLOGI Sessions
     - WWNN 21:00:00:24:ff:55:f5:22 WWPN 21:00:00:24:ff:55:f5:22 fc_id 0x000000
       - LUN's configured (SCSI Type, Version, Vendor, Serial No.)
           LUN ID : 0x0000000000000000  access failure
       - REPORT LUNs Query Response
           LUN ID : 0x0000000000000000
           LUN ID : 0x0001000000000000
           LUN ID : 0x0002000000000000
           LUN ID : 0x0003000000000000
  - Nameserver Query Response
     - WWPN : 21:00:00:24:ff:55:f5:22
     - WWPN : 20:00:00:25:b5:aa:00:02
     - WWPN : 20:00:00:25:b5:aa:00:03
     - WWPN : 20:00:00:25:b5:aa:00:04
     - WWPN : 20:00:00:25:b5:aa:00:05
     - WWPN : 20:00:00:25:b5:aa:00:06
     - WWPN : 20:00:00:25:b5:aa:00:07
     - WWPN : 20:00:00:25:b5:aa:00:08

vnic : 18 lifid: 7
  - FLOGI State : flogi est (fc_id 0x070141)
  - PLOGI Sessions
     - WWNN 21:00:00:24:ff:55:f5:23 WWPN 21:00:00:24:ff:55:f5:23 fc_id 0x000000
       - LUN's configured (SCSI Type, Version, Vendor, Serial No.)
           LUN ID : 0x0000000000000000  access failure
       - REPORT LUNs Query Response
           LUN ID : 0x0000000000000000
  - Nameserver Query Response
     - WWPN : 21:00:00:24:ff:55:f5:23
     - WWPN : 20:00:00:25:b5:bb:00:02
     - WWPN : 20:00:00:25:b5:bb:00:03
     - WWPN : 20:00:00:25:b5:bb:00:04
     - WWPN : 20:00:00:25:b5:bb:00:05
     - WWPN : 20:00:00:25:b5:bb:00:06
     - WWPN : 20:00:00:25:b5:bb:00:07
     - WWPN : 20:00:00:25:b5:bb:00:08

Also, flogi database. Both, initiator and target are logged (VSAN 100, fabric-a):

sh flogi database | i 20:00:00:25:b5:aa:00:01|21:00:00:24:ff:55:f5:22
fc2/15           100   0x1e0000  21:00:00:24:ff:55:f5:22 20:00:00:24:ff:55:f5:22
fc2/16           100   0x1e0141  20:00:00:25:b5:aa:00:01 20:ab:00:25:b5:00:00:00

sh fcns database | i 20:00:00:25:b5:aa:00:01|21:00:00:24:ff:55:f5:22
0x1e0000    N     21:00:00:24:ff:55:f5:22                 scsi-fcp:target 
0x1e0141    N     20:00:00:25:b5:aa:00:01

And default zone settings for VSAN 100 (fabric A):

show zone policy vsan 100
Vsan: 100
   Default-zone: permit
   Distribute: active only
   Broadcast: disable
   Merge control: allow
   Generic Service: read-write

Any comment or though will be appreciated

Rick1776 · ‎11-08-2017

Are all the blades using the same service profile?

Mich El · ‎11-08-2017

Hi Rick,

Thanks for posting. Nope, there are 8 different service profiles. However, in this specific service profile (for 1 of the blades), the boot policy looks good as well as expected wwpns are assigned. And this is also reflected on the adapter output - correct target wwpns are queried, and correct initiator wwpn is logged to the fabric.

Rick1776 · ‎11-08-2017

Try to pin the ports to FI B and see if it will boot up. If not you may have to call TAC to RMA the server blade.