I've had a lovely day of witnessing some trapd.conf and netview rulset
corruption. I believe some parsing errors in trapd.conf (reported
after running addtrap commands) may have contributed to the ruleset
corruption?
This has made me ponderous about causes and solution to these issues.
For instance, when trapd.conf gets corrupted, what commands can you
run to diagnose the errors, relevant logfiles, tools to fix (text
editor)? When trapd.conf is changed, how are the dependent processes
in netview notified (and/or what daemons need to be recycled?).
I knew I had a problem when I loaded the offending ruleset into a
dynamic workspace and nvcorrd promptly ate up all available memory
then crashed.
I'm using AIX 4.3.3, NV 6.0 on this particular server, if that
impacts these questions.
--
|