IBM RS/6000 SP, Problem Determination Manual

Page: 220 / 316

This soft copy for use by IBM employees only.
These files define how the nodes will communicate. The topology files
define the links between nodes, and the clock files define how the clock is
sourced into each switch board and adapter. If your topology or clock file is
wrong, you will not be able to use the switch. Always check that the SDR
contains the right files.
• Hardware
Because the switch is a very stable SP component, hardware problems are
in the common sources of switch-related problems.
7.4 Isolating System Partitioning Problems
As we discussed in Chapter 6, “Error Logging” on page 159, many PSSP
components get partitioned when a system partition is applied. Nevertheless,
many components remain systemwide.
Figure 80. System Partitioning Problems
System partitioning problems usually arise from two sides:
•
Problems applying system partitions
There several ways to recover a failed system partition apply. These ways
usually consist of recovering the SDR (previously saved). If that fails, you
still have a chance to recover your system by executing some SDR
commands that were described in Chapter 5, “System Partitioning” on
page 125.
•
Problems accessing partitions
Usually these problems are related to missing topology files or problems
with the SDR. Remember that several SP components also get partitioned
when you apply a system partition, so be careful to check that your
commands are querying to the right set of daemons.
200
SP PD Guide