Steps to reproduce:

1. Create standard Jira media
2. Create new action with Operation and recovery operation with notify all involved
3. Fire up some trigger which match action
4. Check new Jira ticket in Jira and in Zabbix (both site it be fine)
5. Close problem by button in Update action or by trigger it self
6. Your zabbix instantly die. He is not be able to start again, after cople of second or run in my cas 3-5s he died. I try to increase HA Fail-over delay to 5 minutes but it not help.
7. Only temporary solution is delete (not disable) action which is used for notification.

Result:
Zabbix is restarted every couple of seconds
Expected:
Ticket is updated and server not crash