We have a bit of a unique situation that I don't seem to find any documentation about, and I'm wondering if it can even be done. We have what we call build servers (vm's), that are used to compile and test code, after which the servers are deleted. The developers and systems engineers want to be able to capture performance metrics for these hosts and have, as part of the build script for the vm's, the install logic for the Zabbix agent (for Linux, Windows and Mac). The hosts register in Zabbix successfully and metrics are collected for the entire time that they are running. But after the vm's are deleted, the host obviously is shown as unavailable. I'd like to find a way to capture and keep the metrics in some form (csv, json etc) and then delete the host from Zabbix so that they aren't cluttering up the environment and the queue (i know the queue is just a logical thing). Soooo.....
First, is there a way to export metrics to csv or json or what ever form?
Second, is there a way to automatically delete a host after it becomes unresponsive after a certain amount of time?
Is there any better way to handle this?
First, is there a way to export metrics to csv or json or what ever form?
Second, is there a way to automatically delete a host after it becomes unresponsive after a certain amount of time?
Is there any better way to handle this?