Red Hat® Insights gives your business the ability to predict and prevent problems before they occur. You get real-time, in-depth analysis of your Red Hat infrastructure to proactively identify threats to security, performance, and stability.

My part in this:

I wrote Insight rules for Satellite Performance, rules are written to Tune Satellite in a way so that customers can use Insights and will get the solution to their problem without reporting it as a bug. Most of the time customers using Satellite report bugs which aren't exactly bugs but just an untuned installation of Satellite. Such as using System memory far more than required or using it lower than needed. For such cases, a person should go through the Satellite Tuning guide which we expect he will not :P. So, for that if a person is using Insights then we can help him to know the problem and also the solution to that problem. It works by using insights-core which contains the parsers and combiners which are required in writing rules so as to get the information from the sos-report. Every rule consists of logic through which it decides whether it needs to report or not. Rules are written in python and it is the developers logic which decides what to do and how to do. We just have a problem such as - "The file type should not be nfs for a certain tool version to be installed on that system". So, for that we know that first we need to look for the current version of the tool in the sos-report which is simply a system report of the system on which that tool is installed. A sos-report can be created by developer itself and it stores all the system's information in the form of individual files. For looking up something in that file we have insights-core. There are lots of parsers present at the moment which can be used to write rules according to developers need. Insights is very easy to use for customers but the same is not the case for a developer while writing rules. It's a really complex and tough job to write rules. And the worst nightmare to every developer are the test cases. Writing test cases for the rules is not just writing a simple pytest. It is a bit typical as these are not direct test cases written for a simple python code. Things that a rule should report - ERROR-KEY, Tuning guide link, Conflicting values such as low number of cpu-cores, low version of some software, etc.

To know more visit Red Hat Insights