Chapter 3. Troubleshooting

This chapter provides troubleshooting information for the SGI 1100 server. The following topics are covered:

Error Beep Definitions

Table 3-1 describes the SGI 1100 server error beep definitions. In the beep code sequences, “2” indicates a long beep, and “1” indicates a short beep.

Table 3-1. Error Beep Definitions

Beep Code

Error Description

2-1-2-2

DMA internal register test failure.

2-1-1-1

KBC (keyboard controller) self-test failure or not installed.

2-1-2-1

System DRAM refresh cycle out of specification.

2-1-1-2

Shutdown byte in RTC CMOS R/W failure.

2-2-2-1

 First 640KB of base memory R/W test failure.

2-2-1-2

BIOS code in flash ROM is corrupted.

2-1-1

Video buffer failure.

1-1

Console does not exist and message displays to terminal.

2-2-1-1-1

No memory installed.

2-2

SPD revision is prior to Rev. 1.1 or unavailable for 100-MHz memory DIMM.

2-2-2-1-1-1

Unregistered memory DIMM.


Power-On Self-Test (POST) Checkpoints List

The Power-On Self-Test (POST) is a BIOS procedure that boots the system, initializes and diagnoses the system components, and controls the operation of the power-on password option. When POST executes a task, it uses a series of preset numbers called checkpoints to be latched at port 80h, indicating the stages it is currently running. This latch can be read and shown on a debug board.

Table 3-2 describes the common tasks carried out by POST. Each task is denoted by a unique checkpoint number. For other unique checkpoint numbers that are not listed in the table, see the corresponding product service guide.

Table 3-2. POST Checkpoints List

Checkpoint

Descriptions

01H

Initializes chipset point (I)

02H

Determines BSP and counts total number of CPUs

04H

Determines if the current booting procedure is from cold boot (press reset button or turn the system on), from warm boot (press Ctrl-Alt-Del), or from exiting BIOS setup.

At the beginning of POST, port 64 bit 2 (8042 system flag) is read to determine whether this POST is caused by a cold or warm boot. If it is a cold boot, a complete POST is performed. If it is a warm boot, the chip initialization and memory test is eliminated from the POST routine.

08H

Disables Non-Maskable Interrupt (NMI), Alarm Interrupt Enable (AIE), Periodical Interrupt Enable (PIE), and Update-ended Interrupt Enable (UIE).

These interrupts are disabled in order to avoid the occurrence of problems during the POST routine.

09H

Initializes chipset point (I)

05H

Sets and checks CPU ratio

10H

Tests and initializes DMA controller (8237)

14H

Tests and initializes system timer (8254)

18H

Tests memory refresh; refreshes occurrence verification (IRQ0)

1CH

Verifies CMOS shutdown byte, battery and checksum

Several parts of the POST routine require the system to be in protected mode. When returning to real mode from protected mode, the processor is reset, and therefore POST is re-entered. To prevent re-initialization of the system, POST reads the shutdown code stored in location 0Fh in CMOS RAM. Then it jumps around the initialization procedure to the appropriate entry point. The CMOS shutdown byte verification assures that CMOS 0Fh area is fine to execute POST properly.

Initializes CMOS default setting

Initializes RTC time base

The RTC has an embedded oscillator that generates 32.768 KHz frequency. To initialize RTC time base, turn on this oscillator and set a divisor to 32.768 so that RTC can count time correctly.

1EH

Checks for DRAM and determines DRAM type

2CH

Tests 128K base memory

The 128K base memory area is tested for POST execution. The remaining memory area is tested later.

23H

Initializes BSP local APIC

30H

System shadow RAM

25H

BIOS bootable setting. Check F10 key.

31H

Initialize PCI hot-plug controller

20H

Tests keyboard controller (8041/8042)

Determines keyboard type (AT, XT, or PS/2)

24H

Tests programmable interrupt controller (8259)

Initializes system interrupt

34H

Sizes DRAM

36H

Initializes APIC

3CH

Sets interrupt service for POST

35H

PCI pass 0

4EH

Scans PnP devices

41H

Initializes PCI (1)

51H

Copies module to F segment

50H

Initializes video display

If system has any display card, it should be initialized here via its I/O ROM or corresponding initialization program.

54H

Processes VGA shadow region

4FH

Configures PnP devices

4CH

Checks CPU external frequency

4DH

Records wakeup status

58H

Displays SGI logo (if necessary)

Displays SGI copyright message (if necessary)

Displays BIOS serial number

64H

Tests keyboard interface

The keyboard LEDs should flash once.

5EH

Loads Pentium Pro CPU update code

60H

Initializes SRAM cache capacity

Enables cache function

5CH

Tests memory (except the 128K base memory)

5DH

Builds MP system

8DH

Activates application processor

57H

Initializes hardware monitoring

Checks NVRAM

Initializes IMP card

5AH

Tests SM RAM

5FH

Enables/disables USB function

70H

Initializes parallel port(s)

74H

Initializes serial port(s)

75H

Initializes RDM

78H

Resets math coprocessor

7CH

Checks & initializes pointing device

80H

Sets security status

82H

Builds BCV/IPL table

84H

Initializes keyboard

6CH

Tests and initializes FDD

Note: The FDD LED should flash once and its head should be positioned.

8FH

Builds ACPI tables

95H

IOROM initialization manager

88H

Sets HDD type and features (e.g., transfer speed, mode, etc.)

Tests HDD controller

90H

Displays POST status

96H

IOROM initialization

13H

Re-hooks INT 13h

97H

Moves boot sequence table to F-segment

A0H

Sets time and day

A4H

Initializes security features

A2H

Initializes setup items

A8H

Sets up SMI parameters

ACH

Enables NMI, checks parity if set, and clears screen

AEH

Reserves RDM buffer

B0H

Checks power-on password

Displays configuration mode table

Prepares for booting

B1H

Boots

BDH

Shutdown 5

BEH

Shutdown A

BFH

Shutdown B


POST Error Messages

POST is a BIOS procedure that boots the system, initializes and diagnoses the system components, and controls the operation of the power-on password option. If there is any error during the POST routine, BIOS detects it and shows the corresponding error message on the monitor to guide the technical service engineer on the repair procedure.

Table 3-3 describes the error messages, their possible causes, and corrective actions.

Table 3-3. POST Error Messages

Error Message

Possible Cause

Corrective Action

Memory Error at MMMM:SSSS:OOOOh
(R:xxxxh, W:xxxxh)

DIMMs may be defective.

Replace the DIMMs.

System Management Memory Bad

System Management Memory (SMM) is bad. This may be caused by the malfunction of system green function.

Replace the DIMMs.

 

Battery is Critical Low

For notebook model only.

Recharge the battery.

Keyboard Interface Error

Error in the interface between the system board and the keyboard. The keyboard circuit module may be defective.

Check the keyboard interface circuit or change the keyboard.

Keyboard Error or Keyboard Not Connected

Error in the keyboard, or the keyboard is not connected.

Reconnect or replace the keyboard.

Keyboard Locked

The keyboard lock feature prevents any access to keyboard.

Unlock the keyboard.

Pointing Device Error

The pointing device installed may be bad or the device is improperly connected.

Reconnect or replace the pointing device.

Pointing Device Interface Error

Error in the interface between the system board and the pointing device.

Check the keyboard interface circuit.

Pointing Device IRQ Conflict

The IRQ setting of add-on card and/or system board conflicted with on-board pointing device.

Enter Setup and change the setting of IRQ12.

IDE Drive 0 (1, 2, 3) Error

The IDE drive may be bad, its type may be mismatched, or it was not properly installed.

Check the HDD cable connections and CMOS setup configuration.

Replace the disk drive or the hard disk drive controller.

 

IDE Drive 0 (1, 2, 3) Auto Detection Failed

The hard disk drive may be bad, or it was not compatible with IDE industrial specification.

Check the HDD cable connections and CMOS setup configuration.

Replace the disk drive or the hard disk drive controller.

 

Diskette Drive A Error

Diskette drive A may be bad.

Replace the diskette drive.

 

Diskette Drive Controller Error

This error is caused by any of the following:

1. The diskette drive cable is not plugged into the diskette drive interface on the system board or to the diskette drive.

2. The diskette drive controller is defective.

Check the diskette drive cable and its connections. If the cable is good and properly connected, the diskette drive controller may be the problem.

Change the diskette drive controller or disable the on-board controller by installing another add-on card with a controller.

CPU Clock Mismatch

The user has changed CPU frequency.

When the user changes the CPU frequency, this message is shown once. Then the BIOS adjusts CPU clock automatically.

On-board Serial Port 1 Conflict(s)

On-board Serial Port 2 Conflict(s)

On-board serial port address conflicts with the add-on card serial port.

Change the on-board serial port address in Setup or change the add-on card serial port address.

 

Real Time Clock Error

Real-time clock error.

Check RTC circuit or replace the RTC.

CMOS Battery Bad

CMOS battery power lost.

Replace the on-board lithium battery.

CMOS Checksum Error

CMOS RAM error.

Run Setup again and reconfigure the system.

On-board xxx...Conflict(s)

On Board device resources (e.g., IRQ, DMA, I/O address) conflict.

Try to reassign or disable on-board device resources.

PCI Device Error

PCI device may be bad.

Check the PCI card. Replace if bad.

System Resource Conflict

Some system resources conflict with the resources required by the PCI device.

Run Setup to reconfigure the system.

IRQ Setting Error

Wrong IRQ setting for the PCI device.

Run Setup to reconfigure the system.

Expansion ROM Allocation Fail

The I/O expansion ROM fails to allocate for the PCI device.

Change the I/O expansion ROM address.


Error Message for NMI and Warning Message

A Non-Maskable Interrupt (NMI) causes the CPU routines to be interrupted and the system to be halted

Figure 3-1 describes the NMI error messages, their possible causes, and corrective actions.

Table 3-4. NMI Error Messages and Warning Messages

Error Message

Possible Cause

Corrective Action

RAM Parity Error

DIMM(s) may be defective.

Replace the DIMM(s), or disable parity check in Setup if the model supports it.

I/O Parity Error

The I/O access is not correct.

Check all I/O related circuits (i.e. system I/O controller, memory controller, interrupt controller, DMA controller, etc.)

Press Ctrl_Alt_Esc key to enter SETUP or F1 key to Continue...

A system configuration error is detected, or the hardware configuration does not match the Setup configuration data in CMOS.

Press Ctrl+Alt+Esc to reconfigure the system.

Press ESC to turn off NMI, or any key to reboot

A Non-Maskable Interrupt (NMI) has occurred.

Press Esc to reject NMI error or press any other key to reboot the system.

Insert system diskette and press <Enter> key to reboot

A non-bootable diskette is detected in the diskette drive when the system boots.

Insert a bootable disk in the diskette drive or remove this disk if a hard disk drive is installed.

Equipment Configuration Error

The hardware configuration does not match the Setup configuration data.

Run Setup and reconfigure the system.


IPMI Event Log

In the event of a system crash, System Event Log enables the user to track the error or the abnormal functioning of a component that might have caused the crash.

Table 3-5 describes the type of system event and corrective actions.

Table 3-5. IPMI System Event Log

Type

Description

Action

CPU IERR

IERR is issued when internal CPU operation error occurs. BIOS logs the event and identifies which CPU issues it. See “Jumper and Connector Information” in Appendix A

 for the location of CPU.

Check CPU.

CPU Thermal Trip

Thermtrip is issued when internal CPU temperature is too high. BIOS logs this event and identifies which CPU issues it. See “Jumper and Connector Information” in Appendix A

 for location of CPU.

Restart system.

Check CPU fan connector.

Check CPU fan.

Check CPU.

CPU Processor Disabled

CPU was disabled, as detect result shows abnormal result. It could be CPU thermal issue.

Check CPU fan.

Check CPU.

CPU Temperature

Lower Critical Going Low
Lower Critical Going High

Upper Non-critical Going Low
Upper Non-critical Going High

Upper Critical Going Low
Upper Critical Going High

 

CPU temperature is abnormal. See note for threshold setting.

Restart system.

Check Fan connection.

Check Fan.

Check CPU.

Check System board.

CPU Voltage

Lower Critical Going Low
Lower Critical Going High

Upper Non-critical Going Low
Upper Non-critical Going High

Upper Critical Going Low
Upper Critical Going High

 

CPU voltage is abnormal. See note for threshold setting.

Check CPU.

Check power supply.

Check system board.

CPU Fan

Lower Critical Going Low
Lower Critical Going High

Upper Non-critical Going Low
Upper Non-critical Going High

Upper Critical Going Low
Upper Critical Going High

 

CPU fan is abnormal. See note for threshold setting.

Check CPU fan connector.

Check CPU fan.

System Board Temperature

Lower Critical Going Low
Lower Critical Going High

Upper Non-critical Going Low
Upper Non-critical Going High

Upper Critical Going Low
Upper Critical Going High

 

System board temperature is abnormal. See note for threshold setting.

Restart system.

Check system fans.

Room Temperature

Lower Critical Going Low
Lower Critical Going High

Upper Non-critical Going Low
Upper Non-critical Going High

Upper Critical Going Low
Upper Critical Going High

Room temperature is abnormal. See note for threshold setting.

Adjust room temperature.

System Board Voltage

1.5V/1.8V/2.5V/2.85V for SCSI 1/3.3V/3.3V Standby/5V/2.85V for SCSI 2/12V/-12V/Cache 1/2/Cache 3/4/5V Standby

Lower Critical Going Low
Lower Critical Going High

Upper Non-critical Going Low
Upper Non-Critical Going High

Upper Critical Going Low
Upper Critical Going High

System board voltage is abnormal. See note for threshold setting.

Check power supply.

Remove some devices which are using voltage to reduce system loading.

Check system board.

Keyboard/Mouse Fuse

Current of keyboard/mouse is over the system limit.

Check keyboard/mouse.

USB1/USB2/USB3/USB4 Fuse

 Current of devices connected in USB1/USB2/USB3/USB4 is over the system limit.

Check devices connected in the designated USB port.

Power Supply Predictive Failure

Power supply is dead.

Check power supply.

Chassis Fan Assertion

Chassis fan is dead.

Check chassis fan.

Watchdog BIOS/POST

POST is not completed.

Check BIOS checkpoints list.

Watchdog OS/Load

Problem in loading OS.

Check hard disk.

Watchdog SMS/OS

OS hangs, after loaded.

Check BIOS event log.

Check OS.

Watchdog No Action

System hangs, setting is No Action.

Revise Watchdog settings, if prefer actions automatically carried out by systems.

Watchdog Hard Reset

System hangs, auto Reset.

Revise Watchdog settings, if prefer actions other than Reset automatically carried out by systems.

Watchdog Power Off

System hangs, auto Power Off.

Revise Watchdog settings, if prefer actions other than Power Off automatically carried out by systems.

Watchdog Power Cycle

System hangs, auto Power Cycle.

Revise Watchdog settings, if prefer actions other than Power Cycle automatically carried out by systems.

NVRAM SDR Checksum Error

NVRAM SDR data was damaged.

Rewrite NVRAM.

Replace NVRAM.

NVRAM SEL Checksum Error

NVRAM SEL data was damaged.

Replace NVRAM.

NVRAM FRU Checksum Error

NVRAM FRU data was damaged.

Rewrite NVRAM.

Replace NVRAM.

EMP Remote Login Password Fail

Password error.

Get the correct password.

EMP BMC Disable CPU

BMC disables CPU, after detecting abnormal status of CPU.

Reset the system. If the problem remains, replace CPU.

BIOS Post (Event Data 2)

The event data 2 is POST error code. Setup has to find the POST message table for displaying the message.

Check POST message table.

Secure Model Violation

Unauthorized access.

Follow the correct procedure to access units.

Pre-boot Password Violation-User Password

Incorrect user password.

Get the correct user password.

Pre-boot Password Violation-Setup Password

Incorrect setup password.

Get the correct setup password.

DIMM/RIMM Correctable ECC Error

Memory has ECC (error check and correction) error, but system is able to correct it automatically.

No action needed, but if errors reoccur, check the memory.

DIMM/RIMM Uncorrectable ECC Error

Memory has ECC (error check and correction) error, and system is unable to fix it.

Check memory.

PCI PERR (Parity Error)

Error occurs on PCI-related on-board chipset while doing parity checking. This error message indicates the on-board chipset location which is bus 1 device 0 and function 1. See schematics for location of the chipset.

Check system board.

PCI SERR (System Error)

Error occurs on device or add on card of PCI slot.

Check add-on card.

Check PCI device.

Hard Disk Drive Fault

Errors occur in hard disk drive.

Check HDD.

Drive Backplane Fan Fault Assertion

Errors occur in drive backplane fan.

Check drive backplane fan.



Note: As shown in Figure 3-1, there are four levels for threshold setting: Lower Critical, Lower Non-Critical, Upper Non-Critical, and Upper Critical. If the value is between Lower Non-Critical and Upper Non-Critical, it is normal. If the value is between Lower Non-Critical and Lower Critical or between Upper Non-Critical and Upper-Critical, it is in a warning area.

Figure 3-1. Threshold Setting Levels

Threshold Setting Levels

Index of Error Symptoms

Table 3-6 lists error symptoms and corresponding FRUs and actions.

Table 3-6. Index of Error Symptoms

Error Symptom

FRUs/Actions

CD is not working properly.

Clean the CD.

Clean the optical-head lens.

CD-ROM Drive.

CD-ROM drive tray is not working.

The server must be powered on for the CD-ROM drive tray eject button to function. If the server is on, and the tray does not eject:


Insert the end of a paper clip into the manual tray release opening.

Check CD-ROM drive.

CD-ROM drive is not recognized.

Run BIOS Setup and enable the primary IDE channel.

Check cables and jumpers.

Check for correct device driver.

Diskette drive in-use light stays on, or the system bypasses the diskette drive.

If there is a diskette in the drive, verify the following:

The diskette drive is enabled in the BIOS Setup.

The diskette is good and not damaged. (Try another diskette if you have one.)

The diskette is inserted correctly in the drive.

The diskette contains the necessary files to start the server.

The software program is okay.

If the diskette drive in-use light stays on, or the system continues to bypass the diskette drive, replace the diskette drive.

Monitor problems (general).

Check monitor.

Check display adapter/system board.