There are multiple reasons why there isn't a single metric that shows a host availability. The concept of up or down is relative to which services are important for each individual server for each individual environment.
A simple approach to meet this requirement if a Host is only being monitored with an IC (Infrastructure Cartridge) agent like WindowsAgent or UnixAgentPlus, is to use the monitored metric.
The monitored metric is set to "Monitored" when an IC agent can connect to a host to collect metrics, regardless of what those metrics indicate. Otherwise, it sets this metric to "Unmonitored" or "Unavailable" depending on whether or not the agent is using Ping to validate host availability.
This metric is not a percentage, but this can easily be addressed with a derived metric.
After some time, this metric will be available for custom reporting as in the example below. Just a quick disclaimer, this procedure does not retroactively create metrics, there is currently no supported method to achieve that.
This article explains how to create a custom report similar to the example, but for a different topology type.
© ALL RIGHTS RESERVED. Feedback Terms of Use Privacy Cookie Preference Center