Troubleshooting
89
Are any indications present that show that the malfunction was reported as a memory error, PCI error,
or so forth? The processor now contains the memory controller and PCI Express controller, so faults in
other areas may be attributed to a processor malfunction.
When did the problem occur?
Record exactly when the problem happens (include the date and time).If it happens more than once, keep
a list of all symptoms for each occurrence.
What events preceded the failure?
After which steps does the problem occur?
What has been changed since the time the server was working?
Did you recently add or remove hardware or software?
How long has the server exhibited problem symptoms?
If the problem occurs randomly, what is the duration or frequency?
Observe the server LEDs and their statuses. See the LED Indicators for more information.
Prepare the server for diagnosis
1.
Be sure that the server is in the proper operating environment with adequate power, air conditioning,
and humidity control. For required environmental conditions, see the server documentation.
2.
Record any error messages displayed by the system.
3.
Remove all external peripherals such as the network cables, monitor, keyboard, and mouse.
4.
Collect all tools and utilities necessary to troubleshoot the problem, such as a Torx
screwdriver, loopback adapters, ESD wrist strap, and software utilities.
5.
Determine if the server will troubleshoot offline or online.
6.
To diagnose the server offline, power down the server and peripheral devices. If possible, always
perform an orderly shutdown:
a.
Exit any applications.
b.
Exit the operating system.
c.
Power down the server.
7.
Disconnect any peripheral devices not required for testing (any devices not necessary to power up the
server).
Performing processor procedures in the troubleshooting process
Before performing any troubleshooting steps that involve processors, review the following guidelines:
Be sure that only authorized personnel perform the troubleshooting steps that involve installation,
removal, or replacement of a processor.
Never touch the contacts in the processor socket. THE PINS ON THE SYSTEM BOARD ARE VERY
FRAGILE AND EASILY DAMAGED. If the contacts inside the proces
s
or socket are damaged, you
must replace the system board.
Some processor models require the use of a processor installation tool, and specific steps are
documented to ensure that they do not damage the processor or processor socket on the system
board. For server models that have pins inside the processor socket, remember that THE PINS ON THE