Is there a way to make the checks a little less touchy ?

[Quantum]

What I want to do is to have checks double check their status before going bad.

ie

Standard check = 1 min
Error state is detected so the check goes up to every 1 seconds
is the status is still bad after 10 seconds, then alert.

Is this possible ?

Even an average over time for some of the checks would be ok.

eg, I don't want an alert every time my mail server's CPU spikes to 100% for 3 seconds, I want to know if it's been at 100% for 30 seconds or more.

Thanks

Q

[KS-Soft]

eg, I don't want an alert every time my mail server's CPU spikes to 100% for 3 seconds, I want to know if it's been at 100% for 30 seconds or more.

Its easy, just change "Start when N consecutive Bad results occur" parameter of the action from 1 to 3 (if you perform test every 10 sec, 3*10=30 sec)

Standard check = 1 min
Error state is detected so the check goes up to every 1 seconds
is the status is still bad after 10 seconds, then alert.

Use one "Change time interval" action to reduce time interval when test fails and another "Change time interval" action to restore interval when test returns to good status.
But... If you have hundreds or thousands of tests, I would not recommend to use test interval = 1 sec.

Regards
Alex