My Debian systems have
unattended-upgrades installed, which installs security upgrades automatically, once per day. I also have a Nagios check that reports whether upgrades need to be installed.
In this setup, it is normal that this check can report a failing state, but not for longer than 24h. Can I configure Nagios to consider a service as “up” unless it has been failing for more than 24h?
retry_interval only seems to affect when I get a notification, but I also don't want the front-end to be red during these expected failures.)