MAP4350 Isolating cluster code load counter=2
The CEC enclosure LPAR attempted to IML two times and failed each time. A serviceable event was created. When the LPAR boots, it first loads the AIX operating system, then the functional code and finally the RAS (maintenance package) code. The code load counter is initially set to 0 and is increased by 1 at the start of the code load. If the code load is successful, the counter is reset to 0. If it is unsuccessful, the counter is not reset to 0.
About this task
If the load of the functional code is not successful, the failing cluster creates an AIX error log. A serviceable event cannot be created as the functional code and RAS code were not able to be loaded yet. The partner LPAR on the other CEC enclosure reboots the failing LPAR to attempt to get past the error. If the code load is successful, the code load counter is reset to 0. The AIX error log from the prior unsuccessful attempt will not create a problem as the error was temporary. If the second reboot attempt fails, a final reboot occurs. The AIX code is loaded, the functional code load which would fail is bypassed, and the RAS code is loaded. This leaves the failing LPAR unable to do customer operations, but able to communicate to the management console (S-HMC). The LPAR on the partner CEC creates a serviceable event with an SRC=Bxxx38F0 and uses this MAP for further isolation. That serviceable event does not give the error that caused the code load failures. The failing LPAR should create a separate serviceable event using the AIX error log from the prior unsuccessful attempt. That separate serviceable event should contain repair information for the problem that caused the code load failures.