On 03 Jan 2013 16:06, Scott Schollenberger wrote:
I never could get the MSD and vlogs to tape.
But I restored the operating system and the server completed its IPL
Do not forget to apply the cumulative PTF package again, if the media
used to install from was not the LIC and OS which already had that included.
The problem origin can be found in the joblog of the _spooled_ SCPF
joblog; i.e. not by looking at the active job. The spooled file for the
older SCPF jobs of the failing IPLs are a QPJOBLOG shown in the output
from the request to: WRKSPLF (QSYS *N *N SCPF)
In review of the joblog the assumptions about how one would review a
typical interactive spooled joblog should not be applied; i.e. the last
several messages listed likely will not show the origin. The messages
of interest likely will be even earlier in the joblog, shown sometimes
many pages before the end because all of the messages from the
successful IPL follow the prior failed IPLs. For lack of a
search-backwards feature, the F16=Find on a token like 'Escape' to
locate the likely error is not very ideal. I generally start searching
for all occurrences of CPF9999 or the token *FC, and all message prefix
of MCH to find them.
I personally find value in knowing as much as I can about the failure
regardless that the issue has been circumvented\resolved by a reinstall.