the checks arrive in every minutes, so with 2 minutes interval
(i've tried increase it but it was the same), i dont think so checks
are not fresh.
if i change the service ok and reload i will not get mass set of
alerts. but then revert it and reload again will cause the mass set
of initial unknown state again.
i think in high availability 1 minute is reasonable too, cause the
things can go wrong less than 1 minute.
the interval is not relevant cause check_interval is double of
arrive interval, it can be 5 minutes arrive and 10 minute
check_interval too, but reload cause mass of alerts too.
I am not sure what the problem is, it is working for me.
However, I absolutely assure you a 1 minute check is impossibly
unscalable. I have a
service which monitors a daily backup for example, its check interval
is a little more
than a day. With the backup working and supplying passive updates,
the check is
not inclined to invoke the active dummy command.
i have experience with nagios in high availability enviroment with
1000+ server and passive checks and there was no problem. there the
check command run after freshness threshold reached without enabled
active checks.
it seems to be icinga2 works different.
thank you for your answers and help!
best regards,
attila
_______________________________________________
icinga-users mailing list
[email protected]
https://lists.icinga.org/mailman/listinfo/icinga-users
