IBM RS/6000 SP, Руководство по определению проблем

Страница: 221 / 316

This soft copy for use by IBM employees only.
Chapter 8. Producing a System Dump
There is not much difference in the way system dumps are handled in an SP
environment. The only exception to this is in the primary dump device used by
the nodes when they are running AIX Version 4.1.
•
In AIX Version 3.2.5, the default dump device was /dev/hd7/, a special device
for system dumping.
•
With AIX Version 4.1, the primary dump device was changed to /dev/hd6,
which is the default paging space.
This way, you will not need a special device with additional disk space to
copy a system dump when this is generated.
However, in the SP nodes, the special device /dev/hd7/ is still used, even when
the nodes are running AIX Version 4.1, because the nodes do not have a console
directly attached, and after a system dump is generated, the system intends to
copy it to the /var/adm/ras directory.
Therefore, when you are using the paging space, this system dump must be
copied; otherwise it will get lost when the system starts.
If you do not have space on the /var file system, the system will display a menu
on the console asking for an answer about what to do with the system dump.
This will cause your node to hang while it waits for an answer from the serial
link.
This chapter gives a brief overview of system dumps and provides examples
with commands and scripts you can use to handle them.
8.1 Handling Systems Dumps
The system generates a system dump when a severe error occurs. System
dumps can also be user-initiated by users with
root authority. A system dump
creates a picture of your system
′
s memory contents. System administrators and
programmers can generate a dump and analyze its contents when debugging
new applications.
When the system halts with a flashing 888 followed by a 102, then you know that
a system dump has occurred. When this happens, you must reboot the system
in order to get the system back again.
There are two types of system dumps.
System-initiated Produced if a kernel panic occurs.
User-initiated Generated by a command or by using the key lock and the
reset button.
A system-initiated dump is a snapshot of the kernel. A system dump occurs
when a kernel panic takes place. This dump includes selected kernel structures,
as specified in the dump table. The dump table is a kernel structure containing
a list of structures that have been loaded, such as device drivers and program
code. This table is dynamic, so there is definite rule as to how large the dump
will be. The dump does not include paged memory.

Copyright IBM Corp. 1996
201