Chore: Update Scaler Target Calculation in Grafana Dashboard

[atilsensalduz]

This pull request addresses an issue with the calculation of scaler target values in the Grafana dashboard. Currently, the calculation is done using the rate function over a 5-minute period, but the metric is of type gauge. This can lead to inaccurate target values being displayed on the dashboard.

Provide a description of what has been changed

Checklist

• Commits are signed with Developer Certificate of Origin (DCO - learn more)

Fixes #

Relates to #

[atilsensalduz]

I propose modifying the PromQL expression in the Grafana dashboard to directly use the gauge metric without aggregating it with the rate function over a 5-minute period.

NAME                              REFERENCE                              TARGETS       MINPODS   MAXPODS   REPLICAS   AGE
keda-hpa-scaled-object  StatefulSet/sts       96/90, 9/60   3         8         4          121m

For example, while the HPA indicates the current target value as 96, the dashboard displays a maximum of 80% when calculating the scaled target with the rate function. However, when calculating it without the rate function, the dashboard shows the actual current value that aligns with what the HPA indicates.

sum by(metric) (rate(keda_scaler_metrics_value{exported_namespace=~\"$namespace\", metric=~\"$metric\", scaledObject=\"$scaledObject\"}[5m]))

sum by(metric) (keda_scaler_metrics_value{exported_namespace=~\"$namespace\", metric=~\"$metric\", scaledObject=\"$scaledObject\"})

[atilsensalduz]

Hey @JorTurFer,

Could you help me out with the review process for my PR? Any guidance would be awesome!

Thanks,

[JorTurFer]

Wow, I missed this PR totally 😱
Sorry :( I'll check this today

[JorTurFer]

I've testes the change in my own cluster and the values changes significantly (as it's expected):

Honestly, in my mind the change makes sense totally because the panel's name is Scaler Metric Value so I'd not expect any aggregation watching the results.

[JorTurFer]

/skip-e2e