IBM Power SAS RAID Drive Replacement and Testing

I. Hard Drive Replacement Steps:

1)diag

==> Task Selection

==> RAID Array Manager

==> IBM SAS Disk Array Manager

==> Diagnostics and Recovery Options

==> SCSI and SCSI RAID Hot Plug Manager

==> Replace/Remove a Device Attached to an SCSI Hot Swap Enclosure Device

After selecting the hard drive pdiskn, the following rmdev output is displayed:

┌──────────────┐

│ Running rmdev on pdiskn │

└──────────────┘

When the dialog box showing the hard drive in replace or remove status appears, physically replace the hard drive.

After completion, press Enter.

2)==> Identify a Device Attached to SCSI Hot Swap Enclosure Device.

The list shows this hard drive slot in [populated] status. Do not select the hard drive; during the F3 exit process, you can see configing SAS0 executing.

┌──────────────┐

│ configing SAS0 │

└──────────────┘

Press F3 again to return to the command line state.

Note:

1. Do not use Ctrl+C to interrupt and return to the command line, as this will skip the configing SAS0 step.
2. No need to execute diag's ==> Configure Added/Replaced Devices.

smitty sasdam==>List SAS Disk Array Configuration

The list shows this hard drive slot changing from [populated] to pdiskn.

3)smitty sasdam -> reconstruct a SAS disk array

Select Failed pdiskn; RAID begins rebuild.

Monitor the status until 100% completion.

II. Test Results:

Note:

Only under AIX5.3TL11SP01 did we observe pdiskn becoming 'unknwn' during the rebuild process.

smitty sasdam==> List SAS Disk Array Configuration lists this hard drive slot as [populated], and the original pdiskn hard drive is missing.

III. Unknwn Status Recovery:

cfgmgr –l sissas0   #sissas0 is the SAS bus containing pdiskn.

pdiskn appears normally again; no missing pdisk appeared throughout the entire process.

IV. Conclusion

The above steps reduce the likelihood of the 'unknwn' status appearing after RAID hard drive replacement and provide a possible method to recover the Unknwn status. It is recommended to try cfgmgr –l sissas0 on systems with existing unknwn status, which greatly reduces the number of times system restarts are required for recovery.