Question

Hyperscale disk went offline

  • 27 June 2023
  • 1 reply
  • 318 views

Greetings to All,


I am facing the issue that we have encountered with a disk on our Hyperscale node. Upon checking the system log, we have found that the disk has failed. However, in the ILO, all disks are showing as green.

I would greatly appreciate your help in resolving this issue

/var/log/messages:
*********************
Jun 22 07:59:10 homlab-dth  kernel: sd 2:0:19:0: Power-on or device reset occurred
Jun 22 07:59:10 homlab-dth  kernel: sd 2:0:19:0: [sdw] FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE cmd_age=1s
Jun 22 07:59:10 homlab-dth  kernel: sd 2:0:19:0: [sdw] Sense Key : Not Ready [current] [descriptor] 
Jun 22 07:59:10 homlab-dth  kernel: sd 2:0:19:0: [sdw] Add. Sense: Logical unit not ready, notify (enable spinup) required
Jun 22 07:59:10 homlab-dth  kernel: sd 2:0:19:0: [sdw] CDB: Read(16) 88 00 00 00 00 05 86 03 a8 a8 00 00 00 88 00 00
Jun 22 07:59:10 homlab-dth  kernel: blk_update_request: I/O error, dev sdw, sector 23723223208
Jun 22 07:59:10 homlab-dth  kernel: sd 2:0:19:0: [sdw] FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE cmd_age=1s
Jun 22 07:59:10 homlab-dth  kernel: sd 2:0:19:0: [sdw] Sense Key : Not Ready [current] [descriptor] 
Jun 22 07:59:10 homlab-dth  kernel: sd 2:0:19:0: [sdw] Add. Sense: Logical unit not ready, notify (enable spinup) required
Jun 22 07:59:10 homlab-dth  kernel: sd 2:0:19:0: [sdw] CDB: Read(16) 88 00 00 00 00 05 86 03 a9 b8 00 00 00 f0 00 00
Jun 22 07:59:10 homlab-dth  kernel: blk_update_request: I/O error, dev sdw, sector 23723223480
Jun 22 07:59:10 homlab-dth  kernel: sd 2:0:19:0: [sdw] FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE cmd_age=1s
Jun 22 07:59:10 homlab-dth  kernel: sd 2:0:19:0: [sdw] Sense Key : Not Ready [current] [descriptor] 
Jun 22 07:59:10 homlab-dth  kernel: sd 2:0:19:0: [sdw] Add. Sense: Logical unit not ready, notify (enable spinup) required
Jun 22 07:59:10 homlab-dth  kernel: sd 2:0:19:0: [sdw] CDB: Read(16) 88 00 00 00 00 05 86 03 ac 60 00 00 00 88 00 00
Jun 22 07:59:10 homlab-dth  kernel: blk_update_request: I/O error, dev sdw, sector 23723224160
Jun 22 07:59:10 homlab-dth  kernel: sd 2:0:19:0: [sdw] FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE cmd_age=1s

...

Jun 22 07:59:16 homlab-dth  swatch: calling HP led utils
Jun 22 07:59:16 homlab-dth  swatch: xfs_do_force_shutdown for xfs is ['sdw']
Jun 22 07:59:16 homlab-dth  swatch: ['/hedvig/d23:sdw']
Jun 22 07:59:16 homlab-dth  swatch: swatch mode is nodelay
Jun 22 07:59:16 homlab-dth  swatch: swatchmode is nodelay
Jun 22 07:59:16 homlab-dth  swatch: ['f188ee3b525ad54b5ca$21:/hedvig/d23:sdw']

 

 

 


1 reply

Userlevel 2
Badge +4

Hello,

 

We have had a few experiences with this as well, from hsx point of view the disk is in a pre fail condition.

 

We solved this by logging a case to commvault, i found our old case number 220923-174 that you may choose for referal if you want to.

 

We then got the output we needed logging a case to hpe for disk replacement. And a note: the disk number in commvault commandcenter is not the same as in ilo, you need to copy out the serial number of the bad disk in commandcenter and manualy locate it in ilo to get the correct physical slot on the server itself.

Reply