I'm using Zabbix to monitor a remote network. Access between the Zabbix server and the remote network is via a VPN connection. This VPN connection sometimes terminates and comes back online later (sometimes after 1 minute, sometimes after 1 hour).
Normally when the VPN connection would go down, I would receive 'Unavailable by ICMP Ping' triggers from all hosts, since none of them are reachable by Zabbix at that point.
Recently I have added the VPN interface in Zabbix. When the VPN connection goes down, so does the interface. So I can make all other triggers dependant on the 'VPN Down' trigger. However, once the VPN connection is back up, I still get all those 'Unavailable by ICMP Ping' triggers.
I think it goes like this:
VPN connection down --> trigger created
ICMP Ping check to host #1 failed 1st time --> nothing happens
ICMP Ping check to host #1 failed 2nd time --> nothing happens
ICMP Ping check to host #1 failed 3rd time --> Normally this would create a trigger, but it's dependant on 'VPN connection' trigger. So: nothing happens
ICMP Ping check to host #1 failed 4th time --> nothing happens
VPN connection up again --> trigger solved
Since host #1 has 3 unsuccesful ICMP Ping checks, and the dependancy is now gone, it will create a trigger.
Is there a way to circumvent this? Such as, once the VPN connection is back up, reset all other trigger expressions as if nothing has happened?
Normally when the VPN connection would go down, I would receive 'Unavailable by ICMP Ping' triggers from all hosts, since none of them are reachable by Zabbix at that point.
Recently I have added the VPN interface in Zabbix. When the VPN connection goes down, so does the interface. So I can make all other triggers dependant on the 'VPN Down' trigger. However, once the VPN connection is back up, I still get all those 'Unavailable by ICMP Ping' triggers.
I think it goes like this:
VPN connection down --> trigger created
ICMP Ping check to host #1 failed 1st time --> nothing happens
ICMP Ping check to host #1 failed 2nd time --> nothing happens
ICMP Ping check to host #1 failed 3rd time --> Normally this would create a trigger, but it's dependant on 'VPN connection' trigger. So: nothing happens
ICMP Ping check to host #1 failed 4th time --> nothing happens
VPN connection up again --> trigger solved
Since host #1 has 3 unsuccesful ICMP Ping checks, and the dependancy is now gone, it will create a trigger.
Is there a way to circumvent this? Such as, once the VPN connection is back up, reset all other trigger expressions as if nothing has happened?
Comment