We need a way to support the following workflow:
- Something breaks on Saturday at 5pm. Zabbix sends an alert. (This exists already.)
- Staff receives the alert, triages it, decides it's not truly urgent, and silences that alert until Monday at 9am.
- Sunday at 9am another trigger for the same host generates an alert, which is received and resolved by staff.
- Monday at 9am alerts for the Saturday 5pm event start being sent again.
As I understand it the two possibilities today are:
ACKNOWLEDGE the alert and configure the Action to ignore events which have been acknowledged. But there's no time limit – once acknowledged, an Action is silent forever. If alerts could have a time limit ("ACKNOWLEDGE for 24 hours") that would solve this use case.
Schedule the entire host for maintenance. But then if the problem gets worse or another trigger fires, we won't find out. So if we could schedule a trigger on an individual host for maintenance, this would solve the use case.