Logging & Monitoring

Monitor Cluster Components

Metrics server is a monitoring solution for Kubernetes, every Kubernetes cluster can have only one metrics server, it collects metrics from nodes and pods and places them In-memory (not on disk).

Kubelet has a sub-component called cAdvisor and its responsible for retrieving performance metrics from pods and expose them through kubelet API so they can be used by metrics server.

You can download metrics server by running:

kubectl apply -f https://github.com/kubernetes-sigs/metrics-server/releases/latest/download/components.yaml

After downloading metrics server, view CPU and Memory consumption of nodes and pods:

kubectl top node
kubectl top pod

Lab:

No issues identified.

Managing Application Logs

We can create a logging pod using a definition yaml:

As we create the pod, we can view logs:

kubectl logs -f event-simulator-pod image-processor # -f for realtime

Lab:

No issues identified.

Slides

573KB

Kubernetes+-CKA-+0300+-+Logging-Monitoring.pdf

PDF

Open