Are there universal scripts/libraries for analyzing and identifying non-standard situations in the system?

Question

Algorithms

A question from the category "tell me where to dig?".
1. I'm looking for something that will help track certain system events. For example, Google Analytics has an Alerts service that allows you to send notifications about, in fact, this very excess when a certain (set in percentage) threshold of site visitors rises. There are also a bunch of server monitoring systems that send notifications if a server, virtual machine, or something else has died. For myself, I called such logic "comparison with a fixed reference value."
2. Consider a slightly more complex thing. Imagine that we need to track not reaching a certain threshold, but deviations from a stable state. From my point of view, this is implemented as follows: something (suppose it is a script) receives certain values ​​as input and compares them with the accumulated statistics for the previous time interval. If the current value exceeds the average value for the interval, then an event occurs and the owner of the system is notified about it. This example goes in the column "comparison with a floating reference value".
3. Now even more complex behavior. If in the first two options we worked with known events (and conditionally we can program logic on simple if's), now I want to create events based on the input (and historical) data of the system: analyze the input data, find repeating patterns in them, and create events for which you can already apply the logic described in the first two paragraphs.
Actually, the question is: are there such algorithms/libraries/scripts?
To clarify: I read about data mining, bi, clustering algorithms and everything related to this. I'm more interested in specific products and user experience.

Answer 1

There was a good article here relatively recently on monitoring VoIP telephony, and plotting graphs, try searching in the corresponding blog / tags - the next couple of months.
even a link to the article was found habrahabr.ru/post/132400/

Answer 2

Such systems are tied to technology. But the implementation often falls on your shoulders. Algorithms can be searched for as exception handling - if the system is provided, then everything is fine, and if not, then the editor is in your hands! An example of such D-BUS technology. Can you describe in more detail what product/system you want to monitor?