Instana

1. MicroK8s
2. Minikube
3. kind

I tried all three, in the order above.
What worked for me best was kind, and I will use it for the remainder of this article.

Please note that environment I used is for test/demo purposes only. Kind should not be used for production purposes ("kind was primarily designed for testing Kubernetes itself, but may be used for local development or CI." source)

My setup

• My MacBook Pro 2021, with an M1 chip. Note: This is an ARM architecture, which does make a difference for setting up some of the local Kubernetes infrastructure, e.g. hyperkit is not supported on ARM at the time of writing this article. A colleague of mine tried it on an i7 Mac though and it worked as well
• Podman version 4.2.1
• Kind version 0.16.0

Installing the Kubernetes cluster

Install podman

brew install podman

podman machine init --cpus 6 --memory 12288 --disk-size 50 foo
podman machine start foo
podman system connection default foo

Install KIND

Following the instructions to install kind and setup a cluster with rootless podman - no changes needed from what was described in the related article:

brew install kind 
$ KIND_EXPERIMENTAL_PROVIDER=podman kind create cluster

Once this is complete, you can play around with your freshly deployed k8s cluster. You might have to install kubectl, depending on what is already available on your machine.

isabell$ kubectl cluster-info
Kubernetes control plane is running at https://127.0.0.1:61245
CoreDNS is running at https://127.0.0.1:61245/api/v1/namespaces/kube-system/services/kube-dns:dns/proxy

To further debug and diagnose cluster problems, use 'kubectl cluster-info dump'..

isabell$ kubectl get all
NAME READY STATUS RESTARTS AGE
pod/dnsutils 1/1 Running 46 (53m ago) 46h

NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
service/kubernetes ClusterIP 10.96.0.1 <none> 443/TCP 46h

Installing the Instana agent into the cluster

I went to the "Agents" page of my Instana SaaS instance (More > Agents > Install Instana Agents), and selected YAML based install

I really like the visual guidance towards adding all the required fields.
I then downloaded the yaml (my-agent.yaml)
Before I could deploy it to my cluster, I had to make one minor tweak. In the yaml file, I set one additional property to define initial and max JVM heap size for the Agent.

 containers:
  - name: instana-agent
      image: "icr.io/instana/agent:latest"
      imagePullPolicy: Always
      env:
        - name: INSTANA_AGENT_LEADER_ELECTOR_PORT
          value: "42655"
        - name: INSTANA_ZONE
          value: "isabell-zone"
        [...]
        - name: JAVA_OPTS
          value: "-Xms150m -Xmx512m"

isabell$ kubectl logs instana-agent-hrfm6 -n instana-agent
Defaulted container "instana-agent" out of: instana-agent, leader-elector
Docker socket not found at /var/run/docker.sock
Starting Instana Agent ...
Found 'JAVA_OPTS' settings: '-XX:+ExitOnOutOfMemoryError '.
Skipping the memory calculator: cannot retrieve the maximum amount of memory to be used by the agent from /sys/fs/cgroup/memory/memory.limit_in_bytes .
The memory calculator has not been used, and neither the 'JAVA_MAX_MEM' nor 'JAVA_OPTS' environment variables specify a maximum heap setting; setting the maximum heap to 160m.

isabell$ kubectl apply -f my-agent.yaml

and the cluster shows up as a new K8s cluster under "Infrastructure":

This is now a really nice foundation for a fully local setup to be observed by Instana.