manual start and the monit poll interval

[Michel Marti]

If I manually invoke monit start <service> a short time before monit checks the 
service availability, this leads to unwanted restarts of <service>, since the 
service has not yet been started completely.

Example:

# monit start myservice
Feb 17 09:03:14 myhost daemon.err monit[364]: 'myservice' start: 
/etc/init.d/myservice
Feb 17 09:03:15 myhost daemon.err monit[364]: Monitoring enabled -- service 
myservice
Feb 17 09:03:18 myhost daemon.err monit[362]: 'myservice' failed, cannot open a 
connection to INET[localhost:8080]
Feb 17 09:03:18 myhost daemon.err monit[362]: 'myservice' trying to restart
Feb 17 09:03:18 myhost daemon.err monit[362]: Monitoring disabled -- service 
myservice
Feb 17 09:03:18 myhost daemon.err monit[362]: 'myservice' stop: 
/etc/init.d/myservice
...

Since "myservice" takes some time until it listens on localhost:8080, monit 
restarts it...

How can this prevented? Maybe monit should skip the service test if the 
difference between service start time and service check time is smaller "poll 
interval"?