If you use a relatively complex trigger, like
(system.online |
system.ping |
system.ssh)
And if the system's ssh goes down, Zabbix will only report the value for the first item in this array of checks, instead of the one that actually is having trouble.
So the email will say something like:
"Problem: system.online = ONLINE"
Instead of:
"Problem system.ssh = unavailable"
Even worse, if you look at the "overview" for a host like this, it will show everything that's included in the check as in a "troubled" state, instead of just the troubled item. So the boxes for online, ping, and ssh will all be yellow to reflect a warning trigger just went out.
It means I have to trouble shoot N amount of things for every N thing used in a trigger.
(system.online |
system.ping |
system.ssh)
And if the system's ssh goes down, Zabbix will only report the value for the first item in this array of checks, instead of the one that actually is having trouble.
So the email will say something like:
"Problem: system.online = ONLINE"
Instead of:
"Problem system.ssh = unavailable"
Even worse, if you look at the "overview" for a host like this, it will show everything that's included in the check as in a "troubled" state, instead of just the troubled item. So the boxes for online, ping, and ssh will all be yellow to reflect a warning trigger just went out.
It means I have to trouble shoot N amount of things for every N thing used in a trigger.
Comment