Discussion thread for official Zabbix Template HAProxy

[cova]

Dear all,

Zabbix agent is requesting successfully haproxy status:

26856:20220127:223254.107 Requested [web.page.get["http://172.30.161.49:8404/haproxy?stats"]]
26856:20220127:223254.109 Sending back [HTTP/1.1 200 OK....
[Page stats response]

However in Latest Data for HaProxy I'm just seeing these 4 items:



How can I see results for per example HAProxy {#PXNAME} {#SVNAME}: Number of responses with codes 2xx per second ? I need to setup something else than importing template?

Thanks for your kindly support.

[Taomyn]

Hi, I'd like to propose an additional macro to cleanly specify the HAProxy host to connect to



This continues to use {HOST.CONN} as the default and all direct references replaced with the new macro but then it can be overridden for when it's necessary as in my case when the stats page is only exposed on 127.0.0.1

I made this change myself but when I updated to 6.4 and imported the latest template I forgot and needed to remake them - hence coming here to make the request.

[Peckle]

Hey, firstly thanks for the template.

I noticed for server and backend status items we have a preprocessing to ignore unchanged values every 10 minutes. I understand the idea behind this but in combination with the server/backend down trigger requiring 5 values returning down it means it's takes 50 minutes to trigger an alarm for the back end going down.

This is is a long time when trying to keep services highly available. Wondering if we could remove this processor or add a macro for the number of down messages needed in these alarms so the sensitivity can be dialed.

Thanks in advance