Best practice configuration requirement

[Lwl]

Hi, Does enabling the event function of k8s consume the resources of cluster nodes? Is there a best practice configuration requirement

[Taylor-Sanchez]

Hi @Lwl ,

To answer the best practice configuration, you can find additional information in our documentation here to configure Dynatrace with Kubernetes. Overall, there are different deployment options depending on what fits best for your observability purposes.

https://www.dynatrace.com/support/help/shortlink/installation-k8s

[Mizső · Accepted solution]

Hi @Lwl ,

You can connect to the public API or internal API endponit of Kubernetes. 

Kubernetes API Monitoring | Dynatrace Docs

Event monitoring configuration:

Monitor Kubernetes/OpenShift events | Dynatrace Docs

In case of Internal API end point connection you need a containerized AG (FullStack) instrumentation. In case of AppOnly instrumentation you can also create a containerized AG as a StatefulSet set.

Instrumentation types:

Deployment options on Kubernetes/OpenShift | Dynatrace Docs 

Create manual containerized AG in case of AppOnly injection:

Manually deploy ActiveGate as a StatefulSet | Dynatrace Docs 

For event configuration I always use this without filters (I prefer it, pvc consumes lot of DDU):

Then you can able to create nice dashboards and alerts (there are some perdefined in Settings / Anomaly detection):

I hope it helps.

Best regards,

János