I have a distributed setup where an host whose agent sits behind a satellite has some checks that are in states Ok and Cricical respectively. When I shut down the satellite (i.e. stop icinga2 there) I would expect Icinga Web 2 to reflect some change in the state of the host’s checks, marking them as “stale” or “pending” or similar. When I fix the root cause of a Criticial check and press “Check Now” in Icinga Web 2, its display won’t change to Ok until I bring the satellite up again. This is as expected in my view, because the communication between master and agent has to go via the satellite. But as for the other services “behind” the satellite: Is it not mislieading that they are not clearly marked as (potentially) stale while the satellite is down. Or is my overall view mistaken?
If a checkable doesn’t get checked in time, it will be marked as overdue.
However, if you press check now, the above calculation is reset along with the overdue marker IIRC.
@nilmerg What do you think about the following idea?
DB Web hides the check now button if the check is already scheduled for the past. This way a user still can press check now on a non-OK checkable once, but not more until a new check result arrives. If the latter doesn’t happen in time, the user gets the overdue marker.
Hi Grandmaster (indeed),
Thank you very much. I expected to see the indication in the checks own panel. But what seems to happen is that the check appears under Dashboard / Overdue. That’s fine for me as well.