trigger doesn't respond as I expect

[frater]

I'm running Zabbix 1.8.5 and wanted to add a fairly simple trigger to tell me when the server is unavailable for more than 15 minutes.....

{HOSTNAME} server is down for more than 15 minutes
{Template_Windows:status.min(900)}>0
Disaster

Somehow the trigger goes off immediately together with the normal trigger....

What am I missing??

[Davidus]

You can do

{Template_Windows:status.last(#1)}>0&{Template_Windows:status.last(#2)}>0&{Template_Windows:status.last(#3)}>0

depends on how often you`re doing checks...

[frater]

Thanks, I wasn't aware of that syntax.
I will use it more whenever I use an item where I might change the interval...

But shouldn't the status.min(900)>0 go off after 15 minutes?

I've used this syntax in the past for other items with success.
Why does it trigger together with status.last()#0 ??

[Davidus]

I`m not sure if this is going to work, but try

status.min(0,900)>0

[frater]

I did a reinstall of my Zabbix-server and restored the hosts / templates by using export/import and everything's working as expected now.

I find it hard to rely on Zabbix without testing each item/trigger in real life situations. I therefore don't want to use dependencies yet. The trigger in my first post is now working as expected.

BTW...
If the zabbix process itself crashes you will not really notice it in your dashboard.
I therefore added this 10 minute cronjob into /etc/crontab

*/10 * * * * root netstat -lntp | grep -q zabbix_server || /etc/init.d/zabbix-server start

Sometimes zabbix doesn't start after a reboot (maybe mysql hasn't started yet)?

[Davidus]

do you see errors in your zabbix_server.log file?