trigger worked at 60s but not 21600s?

[cirrhus9.com]

Hello:

On our Zabbix 1.8.10, we have a client that needs a simple method of checking for MySQL (5.x) Replication status and I wasn't overly convinced I needed someone else's perl or bash script to do so, so I created a UserParameter on the target host using

UserParameter=mysql.slave.status,mysql -e "show slave status\G;" | grep Slave_SQL_Running | awk '{print $2}'

and this works as expected.
And double-checking my work from the zbx_server,

/usr/local/bin/zabbix_get -s IP -k mysql.slave.status

I get the Expected 'No' response.

The issue seems to be my trigger. When I created it, it fired an alert every 60s for 5 minutes and I received those with this body:

replication status on db1_zoneC: PROBLEM Last Value: No

Great. Looks like it is correct. Then the Boss says "change the interval to every 2, no 6 hours" so I changed the interval to 21600, this is where the weirdness starts.

We got alerts...
Subject: db1_zoneC MySQL Replication Status: OK - Recovered

MySQL Replication Status on db1_zoneC: OK Last Value: No

Now, I am pretty certain that No means No and it is not "Recovered".

The trigger I am worried about is this:

(({db1_zoneC:mysql.slave.status.regexp(No)})#1)

Attached are the Item, Trigger and Latest Data screenshots.

I have since set the interval back to 300s and have not received any alerts.

/usr/local/bin/zabbix_get -s IP -k mysql.slave.status

still gets the Expected 'No' response.

Nothing has changed on the underlying zabbix server or the mail host that processes the alerts.

I did just now try to change it from Character to Text but this change doesn't return any data and history is now blank. I changed it back and a screen full of Nos.
If you could point out the issue/error, I would be most grateful.

Thank you,

[Colttt]

please try "No"

(({db1_zoneC:mysql.slave.status.regexp("No")})#1)

[cirrhus9.com]

Colttt:

I did as you suggested. No change.

Thank you,No

JJ

Edit: Well, I just got this:
Subject: db1_zoneC 01_MySQL Replication Status: PROBLEM
01_MySQL Replication Status on db1_zoneC: PROBLEM Last Value: Yes

Puzzled.

Edit: So I changed the trigger to

 (({db1_zoneC:mysql.slave.status.regexp("No")})#0)

and we got a Recovered email with "Last Value: Yes" in it.

07/15/2013 03:45:53 PM EDT
seems to be working

I'll give it a couple of days before marking as [SOLVED]