IBM eserver pSeries 690, Service Manual

Page: 813 / 1134

Column 2
(0) Logical processor device number assigned by AIX. You can display these logical device
numbers by issuing the following command on the AIX command line:
lsdev -C | grep proc
Column 3
(3.0) Processor address list used by the service processor.
Column 4
(00) Error status of the processors.
The error status of each processor is indicated by AB , where B indicates the number of errors and A
indicates the type of error according to the following:
1. Bring-up failure
2. Run-time non-recoverable failure
3. Run-time recoverable failure
4. Group integrity failure
5. Non-repeat-gardable error. The resource may be reconfigured on the next boot.
A status of 00 indicates that the CPU has not had any errors logged against it by the service processor.
To enable or disable CPU repeat gard, use menu option 77. The default is enabled.
If CPU repeat gard is disabled, processors that are in the
″
deconfigured by system
″
state will be
reconfigured. These reconfigured processors are then tested during the boot process, and if they pass,
they remain online. If they fail the boot testing, they are deconfigured even though CPU repeat gard is
disabled.
The failure history of each CPU is retained. If a processor with a history of failures is brought back
online by disabling repeat gard, it remains online if it passes testing during the boot process. However, if
CPU repeat gard is enabled, the processor is taken offline again because of its history of failures.
Notes:
1. The processor numbering scheme used by the service processor is different from the numbering
scheme used by AIX. To ensure the correct processor is selected, consult the AIX documentation
before configuring or deconfiguring a processor.
2. The number of processors available to AIX can be determined by issuing the following command on
the AIX command line:
bindprocessor -q
.
v Memory Configuration/Deconfiguration Menu
Enable/Disable Memory Repeat Gard: Memory repeat gard partially or fully deconfigures a memory
book automatically during a system boot. If a memory book has failed BIST POST, caused a machine
check or checkstop, or has reached a threshold of recoverable errors, the memory will remain
deconfigured until repeat gard is disabled or the memory is replaced.
Memory repeat gard is enabled by default.
For more information, see “Configuring and Deconfiguring Processors or Memory” on page 828.
Runtime Recoverable Error Repeat Gard: The runtime recoverable error repeat gard flag controls the
deallocation of the memory if a recoverable error occurs during runtime. If a recoverable memory error
occurs, and the runtime recoverable error repeat gard is disabled, the system will continue running with
no change in the memory configuration. If a recoverable memory error occurs, and runtime recoverable
error repeat gard is enabled, the memory half-book or book in which the error occurred will be taken
offline.
Runtime Recoverable Error Repeat Gard is disabled by default.
These configuration/deconfiguration menus allow the user to change the system memory configuration.
If it is necessary to take one of the memory books partially or completely offline, this menu allows you
to deconfigure a book, and then reconfigure the book at a later time. This menu also allows you to see
if the repeat gard function has partially or completely deconfigured a memory book.
When this option is selected, a menu displays. The following is an example of this menu:
Chapter 7. Using the Service Processor
797