MAP 5900: Hardware boot

MAP 5900: Hardware boot helps you solve problems that are preventing the node from starting its boot sequence.

Before you begin

Note: Use the service assistant GUI if there is no front panel display, for example on the SAN Volume Controller 2145-DH8.

If you are not familiar with these maintenance analysis procedures (MAPs), first read Using the maintenance analysis procedures.

This MAP applies to all IBM Spectrum Virtualize models. However, some models do not have a front panel display; use the service assistant GUI if the node does not have a front panel display. Be sure that you know which model you are using before you start this procedure. To determine which model you are working with, look for the label that identifies the model type on the front of the node.

You might have been sent here for one of the following reasons:

The hardware boot display, shown in Figure 1, is displayed continuously.
Figure 1. Hardware boot display
The node rescue display, shown in Figure 2, is displayed continuously.
Figure 2. Node rescue display
The boot progress is hung and an error is displayed on the front panel
Another MAP sent you here
The node status LED, node fault LED and battery status LED have remained off

About this task

Perform the following steps to allow the node to start its boot sequence:

Procedure

Is the Error LED on the operator-information panel illuminated or flashing?
NO

Go to step 2.

YES

Go to MAP 5800: Light path to resolve the problem.
(From step 1)
If you have just installed the IBM Spectrum Virtualize node or have just replaced a field replaceable unit (FRU) inside the node, perform the following steps:
1. Identify and label all the cables that are attached to the node so that they can be replaced in the same port. Remove the node from the rack and place it on a flat, static-protective surface.
2. Remove the top cover.
3. If you just replaced a FRU, ensure that the FRU is correctly placed and that all connections to the FRU are secure.
4. Ensure that all memory modules are correctly installed and that the latches are fully closed.
5. Ensure that the Fibre Channel adapters are correctly installed.
6. Ensure that the disk drive and its connectors are correctly installed.
7. Ensure that the service controller is correctly installed.
8. Replace the top cover.
9. Place the node in the rack.
10. Turn on the node.
Does the boot operation still hang?
NO

Verify the repair by continuing with MAP 5700: Repair verification.

YES

Go to step 3.
(from step 2)
Check if the system BIOS is reporting any errors. You need to attach a VGA display and USB keyboard to see the BIOS output. The customer should be able to supply a suitable display and keyboard.
1. Turn off the node while ensuring that its data is mirrored and synchronized. See MAP 5350: Powering off a node.
2. Connect the keyboard and the display. Figure 3 shows the location of the keyboard and monitor ports on the 2145-CF8. Figure 4 shows the location of the keyboard and monitor ports on the 2145-CG8. Connect to the VGA port and a USB port on the front or the back of a 2145-DH8.
  Figure 3. Keyboard and monitor ports on the SAN Volume Controller 2145-CF8
  
  Figure 4. Keyboard and monitor ports on the SAN Volume Controller 2145-CG8
  Figure 5. Keyboard and monitor ports on the SAN Volume Controller 2145-DH8, front
  - 3 USB port 1 and 2
  - 4 Video port
  Figure 6. Keyboard and monitor ports on the SAN Volume Controller 2145-DH8, rear
  - 7 USB port 6
  - 8 USB port 5
  - 9 USB port 4
  - 10 USB port 3
  - 11 Serial port
  - 12 Video port
3. Turn on the node.
4. Watch the VGA display.
  - If the POST sequence indicates an error, or if the BIOS Configuration/Setup Utility program indicates an error during startup, you need to resolve the error.
  - If it indicates an error with a specific hardware item, power off the node and remove it from the rack. Ensure the item specified is correctly installed, replace the node in the rack, and then restart the node. If the error is still reported, replace the specified item.
  - If a configuration error is reported, run the Configuration/Setup Utility program option to reset the BIOS to its default (factory) settings.
  - If no boot image can be found on a 2145-DH8 node then contact IBM Remote technical support for help, which could lead to installing the software from a USB flash drive.
5. Turn off the node and remove the keyboard and display.
6. Turn on the node.
Does the boot operation still hang?
NO

Verify the repair by continuing with MAP 5700: Repair verification.

YES

Go to step 4.
(from step 3)
1. Turn off the node while ensuring that its data is mirrored and synchronized. See MAP 5350: Powering off a node.
2. Identify and label all the cables that are attached to the node so that they can be replaced in the same port. Remove the node from the rack and place it on a flat, static-protective surface.
3. Remove the top cover.
4. Remove some of the memory modules.
  - In the SAN Volume Controller 2145-DH8, remove the memory modules in slots 1, 4, 9, and 12, in addition to slots 13, 16, 21 and 24 if the second microprocessor is fitted.
5. Remove all installed adapters.
6. Remove the disk drive, unless this is a 2145-DH8, which can only boot from its boot drives.
7. Replace the top cover.
8. Place the node in the rack.
9. Turn on the node.
Does the boot operation still hang with the booting display (perform the NO action) or has the boot operation progressed (perform the YES action)?
Note: With the FRUs removed, the boot will hang with a different boot failure code.

NO

Go to step 6 to replace the FRUs, one-at-a-time, until the failing FRU is isolated.

YES

Go to step 7
(From step 5)
Remove all hardware except the hardware that is necessary to power up. Continue to add in the FRUs one at a time and power on each time until the original failure is introduced.

Does the boot operation still hang?
NO

Verify the repair by continuing with MAP 5700: Repair verification.

YES

Go to step 7.
(from steps 4 and 6)
1. Turn off the node. See MAP 5350: Powering off a node for more information.
2. Identify and label all the cables that are attached to the node so that they can be replaced in the same port. Remove the node from the rack and place it on a flat, static-protective surface.
3. Remove the top cover.
4. Replace the adapter and the disk drive.
5. Replace the memory modules.
  - In the SAN Volume Controller 2145-DH8, remove the memory modules in slots 1, 4, 9, and 12, in addition to slots 13, 16, 21 and 24 if the second microprocessor is fitted.
6. Replace the top cover.
7. Place the node in the rack.
8. Turn on the node.
Does the boot operation still hang with the booting display (perform the NO action) or does the display progress beyond the initial booting panel (perform the YES action)?
NO

Exchange the failing memory modules for new FRUs and verify the repair by continuing with MAP 5700: Repair verification.

YES
Replace the parts in the following sequence:
- For the SAN Volume Controller 2145-CG8 or 2145-CF8:
  Service controller
  
  System board
Verify the repair by continuing with MAP 5700: Repair verification.