MAP 0270: SCSI RAID problem resolution and verification

Use this MAP to resolve SCSI RAID adapter, cache, or drive problems.

• Step 0270-1
  1. If the system displayed a FRU part number on the screen, use that part number. If there is no FRU part number displayed on the screen, refer to the SRN listing. Record the SRN source code and the failing function codes in the order listed.
  2. Find the failing function codes in the FFC listing, and record the FRU part number and description of each FRU.

  Go to Step 0270-2.

• Step 0270-2

  Is the FRU a RAID drive?

  No

  Go to Step 0270-6.

  Yes

  Go to Step 0270-3.

• Step 0270-3

  If the RAID drive you want to replace is not already in the failed state, then ask the customer to run the PCI SCSI Disk Array Manager using smit to fail the drive that you want to replace. An example of this procedure is:

  1. Log in as root user.
  2. Type smit pdam.
  3. Select Fail a Drive in a PCI SCSI Disk Array.
  4. Select the appropriate disk array by placing the cursor over that array and press Enter.
  5. Select the appropriate drive to fail based on the Channel and ID called out in diagnostics. The Fail a Drive screen will appear.
  6. Verify that you are failing the correct drive by looking at the Channel ID row. Press Enter when verified correct. Press Enter again.
  7. Press F10 and type smit pdam
  8. Select Change/Show PCI SCSI RAID Drive Status > Remove a Failed Drive.
  9. Select the drive that just failed.

  Go to Step 0270-4.

• Step 0270-4
  Replace the RAID drive using the RAID HOT PLUG DEVICES service aid:

  Note: The drive you want to replace must be either a SPARE or FAILED drive. Otherwise, the drive would not be listed as an IDENTIFY AND REMOVE RESOURCES selection within the RAID HOT PLUG DEVICES screen. In that case you must ask the customer to put the drive into FAILED state. For information on putting the drive in a FAILED state, refer the customer to the SAS RAID controller for AIX®.

  1. Select the option RAID HOT PLUG DEVICES within the HOT PLUG TASK under DIAGNOSTIC SERVICE AIDS.
  2. Select the RAID adapter that is connected to the RAID array containing the RAID drive you want to remove, then select COMMIT.
  3. Choose the option IDENTIFY in the IDENTIFY AND REMOVE RESOURCES menu.
  4. Select the physical disk which you want to remove from the RAID array and press Enter. The disk will go into the IDENTIFY state, indicated by a flashing light on the drive.
  5. Verify that it is the physical drive you want to remove, then press Enter.
  6. At the IDENTIFY AND REMOVE RESOURCES menu, choose the option REMOVE and press Enter. A list of the physical disks in the system that may be removed will be displayed.
  7. If the physical disk you want to remove is listed, select it and press Enter. The physical disk will go into the REMOVE state, as indicted by the LED on the drive. If the physical disk you want to remove is not listed, it is not a SPARE or FAILED drive. Ask the customer to put the drive in the FAILED state before you can proceed to remove it. For information on putting the drive in a FAILED state, refer the customer to the SAS RAID controller for AIX.
  8. Refer to the service information for the system unit or enclosure that contains the physical drive for removal and replacement procedures for the following substeps:
     1. Remove the old hot-swap RAID drive.
     2. Install the new hot-swap RAID drive. After the hot-swap drive is in place, press Enter. The drive will exit the REMOVE state, and will go to the NORMAL state after you exit diagnostics.
        Note: There are no elective tests to run on a RAID drive itself under diagnostics (the drives are tested by the RAID adapter).

  Go to Step 0270-5.

• Step 0270-5

  If the RAID did not begin reconstructing automatically, perform the following steps.

  Adding a Disk to the RAID array and Reconstructing:

  Ask the customer to run the PCI SCSI Disk Array Manager using smit. An example of this procedure is:

  1. Log in as root user.
  2. Type smit pdam.
  3. Select Change/Show PCI SCSI RAID Drive Status.
  4. Select Add a Spare Drive.
  5. Select the appropriate adapter.
  6. Select the channel and ID of the drive that was replaced.
  7. Press Enter when verified.
  8. Press F3 until you return to the Change/Show PCI SCSI RAID Drive Status screen.
  9. Select Add a Hot Spare.
  10. Select the drive you just added as a spare.

  If there was no hot spare previously installed in the array, the array will begin reconstructing immediately. Reconstruction time will vary based on the size of the RAID array. Allow 1-2 hours for completion.

  To check the progress of the reconstruction:

  1. Log in as root user.
  2. Type smit pdam.
  3. Select List PCI SCSI RAID Arrays.
  4. Choose the array containing the drive you replaced.

     If the state of the RAID array is reconstructing, then it is in process of reconstructing. If it is optimal, then reconstruction has completed.

  5. Press F10 to exit.

  Go to Step 027017.

• Step 0270-6
  Is the FRU a RAID adapter base card, RAID adapter cache card, or RAID adapter battery?

  No

  Go to Step 0270-15.

  Yes

  Go to Step 0270-7.

• Step 0270-7
  Do you want to change the FRU using a hot-swap operation?

  No

  Power off the system, and remove the RAID adapter. Go to Step 0270-8.

  Yes

  Remove the RAID adapter. Go to Step 0270-8.

• Step 0270-8
  Is the FRU you want to replace a RAID adapter cache card or RAID adapter battery?

  No

  Go to Step 0270-10.

  Yes

  Go to Step 0270-9.

• Step 0270-9

  Replace the FRU onto the existing base card.

  Go to Step 0270-11.

• Step 0270-10

  After physically removing the base card from the system, remove any other good FRUs (RAID cache card or cache battery) from the RAID base card adapter. Plug these FRUs on to the replacement RAID base card adapter FRU.

  Go to Step 0270-11.

• Step 0270-11
  Did you change the FRU using a hot-swap operation?

  No

  Install the RAID adapter assembly into the system. Power on the system and log in to AIX. Go to Step 0270-12.

  Yes

  Install the RAID adapter assembly into the system. Go to Step 0270-12.

• Step 0270-12
  Was the replacement FRU a RAID base card?

  No

  Go to Step 0270-14.

  Yes

  Go to Step 0270-13.

• Step 0270-13
  Attention: Prior to cabling the SCSI RAID adapter to the subsystem, check for preexisting configurations on the replacement SCSI RAID base card. The replacement base card can overwrite your system's configuration data if it already has a configuration written to it. Check it before cabling the SCSI RAID subsystem array.

  Ask to customer to check for preexisting configuration on the SCSI RAID base card. Below is an example of this procedure:

  1. Log in as root (if not already root).
  2. Type smit pdam.
  3. Select List PCI SCSI RAID Arrays.
  4. If no RAID arrays are listed, then there are no preexisting configurations on the base card.
  5. Press F10 key to exit.

  If a preexisting configuration exists on the base card, ask the customer to run the PCI SCSI Disk Array Manager using smitty.

  1. Log in as root (if not already root).
  2. Type smit pdam from the AIX command prompt (if not already in the RAID manager).
  3. Select Recovery Options.
  4. Select Clear PCI SCSI RAID Adapter Configuration. Select the adapter that you just installed. Press Enter to confirm.
  5. Return to the Recovery Options menu (if not already there). Select Resolve PCI SCSI RAID Adapter Configuration. Select Accept Configuration on Drives. Select the adapter that you just installed. Press Enter to confirm. The configuration on the new adapter should now match the configuration existent on the drives.
  6. Press F10 to exit.

  You may now proceed to cable the RAID system array.

  Go to Step 0270-16.

• Step 0270-14

  Ask the customer to resynchronize the RAID array configuration:

  1. Log in as root (if not already root).
  2. Type smit pdam.
  3. Select Recovery Options.
  4. Select Resolve PCI SCSI RAID Adapter Configuration.
  5. Select Retry Current Configuration.
  6. Select the appropriate scraid (SCSI RAID) adapter. A message will be displayed as to the success of the operation.
  7. Press F10 to exit.

  Go to Step 0270-16.

• Step 0270-15

  Other RAID FRUs require that the system be shut down prior to replacement.

  1. If the operating system is running, perform the operating system shutdown procedure (get help if needed).
  2. Turn off the system power.
  3. Replace the FRU indicated by the FFC.

  Go to Step 0270-16.

• Step 0270-16

  Run the diagnostics in system verification mode on the RAID subsystem.

  Did the diagnostics run with no trouble found?

  No

  Go to Step 0270-18.

  Yes

  Go to Step 0270-17.

• Step 0270-17
  1. Use the option Log Repair Action in the TASK SELECTION menu to update the AIX error log. Select scraidX (where X is the RAID adapter number of the RAID subsystem you have been working on).
     Note: On systems with fault indicator LED, this changes the fault indicator LED from the Fault state to the Normal state.
  2. While in diagnostics, go to the FUNCTION SELECTION menu. Select the option Advanced Diagnostics Routines.
  3. When the DIAGNOSTIC MODE SELECTION menu displays, select the option System Verification. Run the diagnostic test on scraidX (where X is the RAID adapter number).
  Did the diagnostics run with no trouble found?

  No

  Go to the Step 0270-18.

  Yes

  If you changed the service processor or network settings, restore the settings to the value they had prior to servicing the system.

  This completes the repair; return the system to the customer. Go to Closing a service call.

• Step 0270-18
  Have you exchanged all the FRUs that correspond to the failing function codes?

  No

  Go to Step 0270-19.

  Yes

  The SRN did not identify the failing FRU. Schedule a time to run diagnostics in service mode. If the same SRN is reported in service mode, go to MAP 0030: Additional problem determination.

• Step 0270-19
  Note: Note: Before proceeding, remove the FRU you just replaced and install the original FRU in its place.

  Use the next FRU on the list and go to Step 0270-2.