User Service Guide, Fifth Edition - HP Integrity rx8640 SEU

Table 4-13 Utility Troubleshooting
ActionSymptom
When troubleshooting utilities, LED indications, and the Error Log
are your most useful tools. Utility problems are often isolated to an
access, power, or firmware problem.
Verify the username & password are correct.Verify the MP port is
not disabled.Verify via Utility LEDs that the MP has heartbeat…if
there is no heartbeat, but power is present, reset the MP by reseating
PCA does not affect the operating system.Verify cable and access
configurations are correct.
Common power failure indications:- Cannot
log into MP.- MP not responding.
Interpret the activity & error logs on the MP. Verify LED settings on
utility PCAs. Interpret the chassis logs.
A system alert at the console indicates a utility
subsystem problem.There is an error in the
chassis log and the Attention LED is flashing.
I/O Subsystem Failures
Troubleshooting I/O faults starts with knowing what firmware (F/W) or software (S/W) was
running at the time of the fault and interpreting the fault's error information in the Error Log.
If an EMS event message was generated, the problem/cause/action statement gives you a good
starting point in troubleshooting the problem. If HP-UX messages indicate some sort of problem,
examine the I/O error logs. Use the log tool as the starting point for troubleshooting these
problems.
I/O faults gather error information from different FRUs. For example:
- HPMC or MCA F/W saves all logged fatal and uncorrectable error info from CC, SBA, LBA,
and HBA CSRs.- HP-UX & EMS S/W generates event messages from H/W correctable faults from
memory DIMMs, CPU caches, and HBA functions and their external devices. - ISL & ODE S/W
provides console messages with generic error information about CPU, MEM, I/O, and device
FRU function failures and loop / halt until the partition is hard reset.- BCH or EFI F/W also
provides console messages about failed CPU, MEM, and console operations, while early self-test
F/W relies on LEDs and chassis codes to convey error information about failed cell and chassis
FRUs.
Table 4-14 I/O Problems
ActionSymptom
If an EMS event message was generated, the problem/cause/action statement
gives you a good starting point in troubleshooting the problem. If HP-UX
messages indicate some sort of problem, examine the I/O error logs. Use
the log tool as the starting point for troubleshooting these problems.Examine
the console logs for HP-UX messages indicating problems with the I/O
system. Look for EMS event messages indicating I/O system problems.
Examine the I/O error logs. Use the MP’s "io" & "ps" commands to figure
out which cell is connected to each individual I/O card cage.
HP-UX or EMS messages indicate I/O
problems.
Booting Failures
Booting problems are one of the most difficult problems to troubleshoot because of the many,
many types of failures that can prevent the server from booting. Your initial troubleshooting
approach should be based upon the server environment found when you begin. For instance, if
you have just added a new cell, updated microcode, or taken some other invasive process
recently…then reviewing those actions more closely are a logical first step.
If the server was running normally and was forced into an unsuccessful reboot by a power failure
- then verification of server power and a review of the Error Log would be in order. Additionally,
recent customer activities may suggest a possible troubleshooting path to pursue.
92 Troubleshooting