Creat automated data center shutdown in case of thermal issue

Hello ,

I am looking into a process to automate data center shut down or part of it ,in case of room cooling system malfunction and temperature rais to critical condition .
My idea is to create icinga2 command event using ncp external command that execute simple shut down command script , and calling it using ncp api, on each rack we are monitoring the ambient servers/switch temprature sensor .
The ncp part tested and working , then i started to plan how to do the logic behind data center shutdown, if only part of the data center is having cooling issue , or only few sensor/s having temperature raised above threshold ?, so i started thinking of dividing the data center to racks and create business process to each rack and combine it on big data-center , so i would have controle on each rack /part of the data center or all of it
My question is given above scenario , is it possible to accomplish it with business process , along with command event?
Please advice

It shoud work, as you can create a normal check against businessprocess status, but I would at least thin 3-4 times again, before creating some process which can shutdown my datacenter automatically. Esspecially with icinga, which ist not build for 100% failsafe check results. There are far to many points where wrong informations can shutdown your datacenter.

1 Like

Thank you for the answer, what I don’t understand is the process for BI and the alerting /command event attached together,
If I am doing per rack for example min 3 , how it triggers the command?

You can use the icinga2 cli command to check if a business process is critical and use this for a check as comand. See f.e.

icingacli businessprocess process check --help

for teh checkcommand look at this file:

1 Like