[monit-dev] 4.10.1 too fast for its own good....

[Aaron Scamehorn]

Hello,
 
We've first started experiencing this problem in 4.9, and it continues
in 4.10.1:
 
We use monit to start & stop our processes daily.  Every day at 6am we
start them, and every day at 11pm we stop them.

Occasionally, at startup, monit will start a process, then in the same
second, start it again (thinking the first invocation failed).  Below is
the syslog output:
 
Dec 19 07:05:01 wolverine monit[18493]: start service 'trdClientCap' on
user request
Dec 19 07:05:01 wolverine monit[18493]: 'trdClientCap' start:
/cogcap/ccts/bin/trdClientCap
Dec 19 07:05:01 wolverine monit[18493]: 'trdClientCap' process is not
running
Dec 19 07:05:01 wolverine monit[18493]: 'trdClientCap' trying to restart
Dec 19 07:05:01 wolverine monit[18493]: 'trdClientCap' stop:
/cogcap/ccts/bin/trdClientCap
Dec 19 07:05:02 wolverine monit[18493]: 'trdClientCap' start:
/cogcap/ccts/bin/trdClientCap
Dec 19 07:05:11 wolverine monit[18493]: 'trdClientCap' failed to start
Dec 19 07:05:12 wolverine monit[18493]: 'trdClientCap' process is not
running
Dec 19 07:05:12 wolverine monit[18493]: 'trdClientCap' trying to restart
Dec 19 07:05:12 wolverine monit[18493]: 'trdClientCap' start:
/cogcap/ccts/bin/trdClientCap
Dec 19 07:05:12 wolverine monit[18493]: 'trdClientCap' failed to start
Dec 19 07:05:13 wolverine monit[18493]: 'trdClientCap' started
Dec 19 07:05:22 wolverine monit[18493]: 'trdClientCap' process is
running with pid 24708


I believe the first occurance of the "process is not running" and
"trying to restart" are the problem.  Monit is not even waiting 1 second
to determine if startup failed.

The end result is that I now have 2 processes running.


Thanks,
Aaron