After deploying the Self-Monitoring policies on a Solaris agent v11.13.007,
when the agent starts warning alerts are shown.
Warning opcmona (Monitor Agent) "An error occurred in the processing of the policy 'OA-SelfMonTstAll'. Please check the following errors and take corrective actions. (OpC30-797) Monitor program 'OA_selfmon_run OA-SelfMonTstAll' of source Source failed. Result=1. Output of monitor program: /var/opt/OV/bin/instrumentation/OA_selfmon_run: LC_ALL=C: is not an identifier (OpC30-732)"
Warning opcmona (Monitor Agent) "An error occurred in the processing of the policy 'OA-SelfMonVerifyMon'. Please check the following errors and take corrective actions. (OpC30-797) Monitor program 'OA_selfmon_verify 180 OA-SelfMonVerifyMon' of source Source failed. Result=1. Output of monitor program: /var/opt/OV/bin/instrumentation/OA_selfmon_verify: LC_ALL=C: is not an identifier (OpC30-732)"
Problematic scripts are (on managed node):
On mgmt srv (in instrum dir, after agent registration):
Which is part of HPOpsAgt Category that is shipped with OA.
For example if script is executed in cmdline on the agent, it fails like:
./OA_selfmon_verify 180 OA-SelfMonVerifyMon
./OA_selfmon_verify: LC_ALL=C: is not an identifier
Modify both scripts on mgmt srv side (see below) and redistribute instrumentation to such Solaris agent:
change the two scripts on management server:
The two lines below:
to the following:
Then redeploy the policies.
After doing this, no warning messages occur and SelfMon functionality should work fine.
This is documented in the following KCS document: http://support.openview.hp.com/selfsolve/document/KM01001122