my action didn't execute when the trigger is TRUE

[lth0721]

a create a sh script named a.sh

#!/bin/bash
echo "$@">/home/zabbix/a.txt

and then add a media type echo

and also add the media echo for user test

then add a trigger for host 10.0.1.8

the add the action test2 for host 10.0.1.8

but when the smtp server down on 10.0.1.8,the action didn't execute...

help me..

[zabbix_zen]

Hi.

When using Recovery messages it's mandatory to add the Action condition:
Trigger value = "PROBLEM"

Try unchecking 'Recovery message'

Does your file have the correct permissions?

[lth0721]

thanks for you reply，zen.
I unchecked the 'Recovery message'
and the file also have rigth privileges
but unfortunately the action still didn't execute
my zabbix version is 1.6.6,is that a bug?

[zabbix@localhost ~]$ ll /home/zabbix/a.txt
-rw-r--r-- 1 zabbix zabbix 0 10-29 09:02 /home/zabbix/a.txt

[zabbix@localhost ~]$ ll /home/zabbix/bin/a.sh
-rwxrwxr-x 1 zabbix zabbix 41 10-28 14:08 /home/zabbix/bin/a.sh

[lth0721]

by the way
I only find this info in /tmp/zabbix_server.log,is there something wrong with my set,,I think there should be more info..
6375:20091029:055454 Executing housekeeper
6375:20091029:055457 Deleted 0 records from history and trends
6375:20091029:065557 Executing housekeeper
6375:20091029:065600 Deleted 0 records from history and trends
6375:20091029:075700 Executing housekeeper
6375:20091029:075703 Deleted 0 records from history and trends
6375:20091029:085803 Executing housekeeper
6375:20091029:085805 Deleted 0 records from history and trends
6375:20091029:095905 Executing housekeeper
6375:20091029:095908 Deleted 0 records from history and trends

and I set a trigger to monitor my log file also didn't run..beacuse I can't find any log in agent log file about it..

I am a newbie to zabbix..if there have some special rules in confoguration..pls tell me

[zabbix_zen]

"Deleted 0 records from history and trends"

simply means your Zabbix deploy is young enough to don't have any Item's data to throw away, as configured in your Item's Keep history/trends (in days)

I'm not understanding what you're trying to accomplish with your Trigger, did you already tested the Linux templates+Triggers that come in a vanilla Zabbix Server+Agent setup?

Though it has nothing to do with your problem, your Server has the default value for housekeeping and is performing it hourly, you can change in zabbix_server.conf
'HousekeepingFrequency' to a little bigger value.

[lth0721]

thanks。。again

but in the item had set the keep time as the pictures..
it seams it didn't save the data in db...

[zabbix_zen]

If in Monitoring -> Latest Data you aren't receiving Latest value for 'Number of running processes apache' in the Processes List then something's wrong with your zabbix_agentd.

Check zabbix_agentd.log for errors

[lth0721]

the item is running and receive rigth data from agentd but the action based on this trigger is not run..,
for example I set a trigger like {10.0.1.8roc.num[httpd].last(0)}=0
but when the number change to 0,the action didn't execute...and don't find any log about it..

[zabbix_zen]

Well if that's it so,

then it's probably on the 'Send message to user test' part.

Does user Test belongs to a Zabbix User Group with Read Permissions to the Host/Host Group you're trying to monitor?

[pietux]

I call it as bug

This sounds exactly like the bug me and some others have encountered.
Zabbix simply does not send any alerts for tripped trigger events.

Please see thread: http://www.zabbix.com/forum/showthread.php?t=13769

Petri

[lth0721]

Well if that's it so,

then it's probably on the 'Send message to user test' part.

Does user Test belongs to a Zabbix User Group with Read Permissions to the Host/Host Group you're trying to monitor?

thanks you wake me up....
that's the problem location..many thanks for your so much help to me..sorry to trouble these days..

[zabbix_zen]

I don't see enough data anywhere to call it a bug.

[bashman]

Hi,

I want to run two remote commands, one on a problem and another on recovery.

I have tested the first one (Trigger value = "PROBLEM") and it works fine, but I haven't tested the second.

Can I use Trigger value = "OK" for the recovery command?.

[zabbix_zen]

Affirmative

[bashman]

Ok, thanks. I'll test and post my results.