Environment having hosts with many Zabbix agent (active) items.
Maintenance mode having no data collection.
(See the attached screenshot.)
Maintenance starts 23:00
Hosts are powered down right after that.
Hosts are powered up at approx. 06:00
Maintenance ends 06:30
PROBLEM: Items get into Zabbix queue until the scheduled next check causing huge queue spike (thus firing the configured huge queue trigger).
The queue drops as soon as the next scheduled item check occurs.
Suggested possible solution: Not to queue agent active items after maintenance until scheduled next check.