Hfi; Hfi Health Check - IBM Power Systems 775 Manual

The LNMC topology of the specified cage or FSP must be changed from 128D to 8D. To make
the change, the cage must be Standby (FSP powered on with LNMC running, and the Power
775 not powered on). Complete the following steps to update the topology on the cage:
1. Check that frame 6 cage 11 is at Standby by powering off the cage (if needed) and then
stopping CNM by using the following command:
# chnwm -d
2. If the topology in the Cluster DB and CNM must be updated, the chdef command must be
issued before CNM is restarted. If the topology is correct in the Cluster DB and CNM,
restart CNM and then update the topology to LNMC by issuing the following
chnwsvrconfig command:
# chnwm -a
# chnwsvrconfig -f 6 -c 11
3. Power on the cage and confirm that all topologies show the wanted values by issuing the
following commands: lsnwtopo, lsnwtopo -C and lsnwtopo -A (specifically in this case:
lsnwtopo -f 6 -c 11):
# lsnwtopo
ISR network topology that is specified by cluster configuration data is 8D:
# lsnwtopo -C
ISR network topology in use by CNM: 8D:
# lsnwtopo -f 6 -c 11
Frame 6 Cage 11: Topology 8D, Supernode 5, Drawer 0
Important: CNM must be stopped before the topology updates are made to the cluster DB
via the chdef command and then restarted so that CNM picks up the change. In addition, if
the chnwsvrconfig command is used to update LNMC on an FSP, the cage must be
powered off (Standby) before stopping CNM, restarting CNM, and issuing chnwsvrconfig.
The drawer is powered on, IPLed after CNM is restarted, and chnwsvrconfig is issued.

4.3 HFI

The section describes an overview of some of the commands, tools, and diagnostic tests that
are used for identifying and resolving HFI-related problems.

4.3.1 HFI health check

It is a good idea to run an HFI health check to identify potential issues within the HFI
infrastructure; for example, after a service window or scheduled power cycle of the cluster.
Some examples of the lsnwdownhw and lsnwlinkinfo commands are presented in the section
to determine whether more problem troubleshooting and link diagnostic tests are needed
relative to the HFI network. A certain level of intuition and knowledge of the specific cluster
hardware configuration is needed when the output of the lsnwdownhw and lsnwlinkinfo
commands is interpreted.
Chapter 4. Troubleshooting problems
259