Finding the source of scenario failures

[jterhune]

I just finished setting up web monitoring on our new Zabbix install and I am having a problem with what I believe to be false positives.

It is very simple monitoring:

One step
No authentication
30 Second update interval
30 Second timeout
Searching for one string of text
Status of 200

I have a trigger set up with:

{Web Server:web.test.fail[site].last(#4)}>0

To my understanding, this means that this trigger should go off if there are 4 consecutive failed tests. I'm running into two issues with this.

One, the web server seems to be failing tests randomly and for no particular reason. Occasionally a single test will fail, and I can't really find out any information as to why. I can't find anywhere in a log as to which condition wasn't met. When I look at "response code", I see 200s across the board, so I have no clue why this is failing.

Secondly, even though these failures are random, single events, I seem to get a notification after every one - not after 4 consecutive. Even the notification I get says the last value was 0.

Does anyone have any advice on tracking down what exactly is causing these failures?

[JBo]

I have a trigger set up with:

{Web Server:web.test.fail[site].last(#4)}>0

To my understanding, this means that this trigger should go off if there are 4 consecutive failed tests. I'm running into two issues with this.

That's wrong, last(#4) means that forth most recent value.
If you want to check that last four values are all greater than 0, you should use:

{Web Server:web.test.fail[site].min(#4)}>0

Regards,
JBo