In order to move to a hard non-ok state, a service can require multiple (i.e. max_check_attempts) unsuccessful runs of the check plugin. Is there a way to require multiple successful plugin runs in order to return to a (hard) ok state? I have printers where, if toner is low (not empty), from time to time, the 1104 entry vanishes from the Printer Alert Table for several seconds, only to return at another index shortly after. As I'm using a combination of SNMP traps and table polling, the trap sent upon removal immediately make my printer-toner service switch to OK, after which it takes some time (one trap sent upon table re-apperance and two polls) to return to WARNING. Even if I remove traps from the equation, my table poll may happen during the some-second period the 1104 is gone.
I do know about flapping, but that's not what I want. I want protection from transient success, like I'm protected from transient failures. _______________________________________________ icinga-users mailing list icinga-users@lists.icinga.org https://lists.icinga.org/mailman/listinfo/icinga-users
