Detection of frequent issues, documenttion question

[henk_stobbe]

Hello,

Frequent issue detection,

In the beginning it it states For details on event severities, see event types.

but in the example it switched to:

So it seems al event types are created equal, but some events also have an extra severity weight?

Does dis apply in all cases?  I assume not available is not available?

So in short, who can confirm that 

Event types have severities

And events can have severities, and these are used in this case.

KR Henk

[marina_pollehn · Accepted solution]

In terms of events severity, the percentages are meant. So, a CPU of 99% represents a more severe event than one with a CPU of 70%. Severity is in this case purely based on the CPU value. Of course you can also have other events, like a response time degradation. Then the severity will be measured in seconds or milliseconds. 

The documentation does not compare slowdowns and error severity levels or something like this here. Severity of the individual events - used for frequent issue detection - is purely the number of the metric.

I am hoping this answers the question a bit - asking and answering definitions is sometimes not so easy in terms of understanding each other. 😉

[marina_pollehn]

Hence, I would answer your question: "So it seems al event types are created equal, but some events also have an extra severity weight?" with no. In terms of frequent error detection, these event types are ''equal'' and each have their own set of durations and severities per event. So we are not comparing CPU events with errors here.

[henk_stobbe]

Hi Marina,

Thanks for your reply, you are right. So de correct sentence would be, most events are created equal within their group, but within a group  events can have an extra weight/severity?

KR Henk