Put newly configured already running ScheduledDowntime immediately in effect

[efuss]

Put running downtimes in effect

If Icinga2 was restarted with a newly configured downtime that should
be in effect at the time of restart, the should-be-running segment of
it was not put into effect.

Add new LegacyTimePeriod::FindRunningSegment() and
ScheduledDowntime::FindRunningSegment() functions, call the latter in
ScheduledDowntime::CreateNextDowntime() before trying the old
ScheduledDowntime::FindNextSegment().
Before that, remove some dead code and a spurious (or so I think)
debug string.

This is (or so I hope) the easier part of Pull Request #6579

[mcktr]

Thanks for splitting up the PR.

Test

object Host "devel-host-001" {
	import "devel-host-template"

	address = "127.0.0.1"

	vars.notification["mail"] = {
		groups = [ "icingaadmins" ]
	}
}

// newly configured already running downtime
object ScheduledDowntime "devel-downtime-001" {
	host_name = "devel-host-001"

	author = "icingaadmin"
	comment = "Some comment"

	ranges = {
		"2018-10-19" = "10:00-11:50"
	}
}

// configured downtime with the beginning in the future 
object ScheduledDowntime "devel-downtime-002" {
	host_name = "devel-host-001"

	author = "icingaadmin"
	comment = "Some comment"

	ranges = {
		"2018-10-19" = "11:55-12:00"
	}
}

• Immediately after starting icinga2 the newly configured and already running ScheduledDowntime is put in effect. The host object is put in downtime and a downtime start notification is send out.
• the second downtime in the future was correct planned and put in effect at 11:55.

This fixes #4790

[dnsmichi]

Thanks. I trust your expertise on this, I won't test it again.

[dnsmichi]

@efuss thanks for the patch 👍