HP -UX B6941-90001 Administrator's Reference Manual page 418

Tuning, Troubleshooting, Security, and Maintenance
Troubleshooting: Specific Problems
MPE/iX Managed Node Runtime Problems and
Solutions
Problem Extremely long time for command broadcasting and application startup.
Description The command broadcasting and application startup are done within jobs.
When the job limit is reached, the jobs are queued. Non-ITO jobs also
increase the number of running/pending jobs. By default, ITO runs one
job to control its agents and up to four additional jobs for command
broadcasting and/or application startup.
Solution Increment the job limit (HPJOBLIMIT) if required.
Problem When distributing command, action, or monitor scripts or programs, it
may happen that current actions, commands, and monitors cannot be
replaced.
Description The commands, actions, or monitors are still in use (scripts or
programs are running; text file is busy). The user receives a warning to
this effect. In most cases, this causes no problems, since the existing
actions, monitors, or commands are not often modified (in other words,
the newly-distributed files are equivalent to those in use).
Solution If the user wants explicitly to change a program or script on MPE/iX
which is running, he or she must: Stop the MPE agents:
opcragt -stop <MPE-NODE>
Repeat the distribution (the distribution will restart the agents again).
418 Chapter 10