Metrics tab in Services app : how does it work ?

Hello,

A Metrics tab in Services app appeared a couple of months ago. I just recently dug into it and I have to admit I don’t know what metrics it shows. I tried to fetch some documentation about it but I failed.

I find it to be more or less a duplicate of Overview tab. Example :

Overview

Metrics

Moreover it seems that this tab doesn’t show the same metrics depending on the context. Here is what it shows for another service :

Can someone help me understand how this tab beheaves ?

Edit : we use SDv1

for some services you can also find this in metrics tab:

rastislav_danis_0-1789373883368.png

link points into this doc: Customize service detection in Service Detection v2 — Dynatrace Docs

Mentioned doc should answer your question, why there are various metrics (or none) in metrics tab of service.

Thanks for your reply @rastislav_danis. I should have pointed it out in my initial message : we use SDv1 and the doc you provide is for SDv2, so it is probably not useful in our context.

It seems that SDv2 is only source of data in this tab. There are some default SDv2 rules enabled even if you don’t use them.

The metrics tab shows additional metrics (on top of the standard ones in the overview), which have the dt.entity.service (current, classic explorer) or the dt.smartscape.service (new explorer) dimension equal to the service. For example metrics ingested from your apps using micrometer or opentelemetry.

Thanks for your reply @Julius_Loman. After further reasearch from our side, it seems we kind of agree :+1:

However, we find it to be a strange design. The interest of the additional metrics we configured is to be able to filter with a dimension. That is why the metrics’ names finish with “by instance”. As we can’t do that in the Metrics tab, these metrics are worthless in this context and can even be misleading. One can wonder “why can’t I filter for one instance whereas the metrics is called ‘Service failure count by instance ?’”.

Admins should be able to configure how metrics appear in this tab.

What is the source of your metrics named “by instance” ? Is what is the metric key? aren’t they custom metric you send to Dynatrace or are scraped?

They are custom metrics via Open Pipeline on spans. One of the keys : span.instance.request.response_time.

If you have custom metrics named similar to the built in ones, they will be in the metrics tab. The core dt.service metrics are on the overview tab.