We're eager to learn about your KEDA deployments!

[tomkerkhove]

As open-source maintainers, we are eager to learn about how you are using KEDA and how big of a deployment you have!

If you are allowed to and want to share, we'd love to learn how many scaled objects/scaled jobs you have in your cluster and what your experience has been!

We are also considering introducing our own performance benchmarking which you can learn more about on #4411.

[wiseelf]

Current amount of ScaledJobs and ScaledObjects.

% k get so -A --no-headers=true | wc -l
    1426
% k get sj -A --no-headers=true | wc -l
    704

Had an issue with kube-api request rate, so I have to tune some keda parameters and I'm using cached metrics as well.
My config for KEDA (maybe the values too high, but it works well)

## Extra KEDA Operator and Metrics Adapter container arguments
extraArgs:
  keda:
    kube-api-qps: 150
    kube-api-burst: 250
  metricsAdapter:
    kube-api-qps: 150
    kube-api-burst: 250

## Extra environment variables that will be passed onto KEDA operator and metrics api service
env:
  - name: KEDA_SCALEDJOB_CTRL_MAX_RECONCILES
    value: "50"
  - name: KEDA_SCALEDOBJECT_CTRL_MAX_RECONCILES
    value: "100"
  - name: KEDA_METRICS_CTRL_MAX_RECONCILES
    value: "50"

For the SO,SJ i do specify useCachedMetrics: true

[wiseelf]

Here is graph showing max(keda_scaler_metrics_value) by (namespace) for the last 1 hour.

[tomkerkhove]

Nice, looks like it's doing fairly good!

[tomkerkhove]

PS: Do you want to become a listed end-user on https://keda.sh/community/?

[wiseelf]

Sure, why not? :) Our company uses KEDA for many customers. I think it would be nice.

[tomkerkhove]

Wonderful, thank you! It is documented on https://github.com/kedacore/keda-docs#become-a-listed-keda-user but it's just a PR away.

I highly appreciate it!