Demo Telepresence Basics

This demo walks through a small example application and shows how Telepresence lets your local machine behave as if it is running inside a Kubernetes cluster. That makes debugging cluster-internal services — using your local editor, debugger, or network tools — fast and convenient without port-forwarding or building special images. Overview

Three application services:
- auth service
- product service (entry point)
- inventory service
Flow: client → products → inventory (products aggregates inventory data and returns it). The auth service runs independently.

Application layout

auth/ — Node.js app (index.js, package.json, Dockerfile)
inventory/ — Node.js app (same structure)
products/ — Node.js app (same structure)
k8s/ — Kubernetes manifests for Deployments and Services

Service responsibilities

Service	Role	Notes
`products-service`	Entry point for user requests	Aggregates product data and calls `inventory-service` for stock counts
`inventory-service`	Provides inventory counts	Cluster-internal
`auth-service`	Authentication mock	Independent service used in examples

Below are the key manifest excerpts and the sequence of steps used in the demo. Deployment and Service manifests Inventory Deployment + Service (example)

apiVersion: apps/v1
kind: Deployment
metadata:
  name: inventory-depl
spec:
  selector:
    matchLabels:
      app: inventory
  template:
    metadata:
      labels:
        app: inventory
    spec:
      containers:
      - name: inventory
        image: sanjeevkt720/telepresence-inventory
        ports:
        - containerPort: 3000
          name: web
---
apiVersion: v1
kind: Service
metadata:
  name: inventory-service
spec:
  selector:
    app: inventory
  ports:
  - port: 3000
    targetPort: 3000

Products Deployment (entry point)

apiVersion: apps/v1
kind: Deployment
metadata:
  name: products-depl
spec:
  selector:
    matchLabels:
      app: products
  template:
    metadata:
      labels:
        app: products
    spec:
      volumes:
      - name: db-config
        configMap:
          name: db-info
      containers:
      - name: products
        image: sanjeevkt720/telepresence-products
        volumeMounts:
        - name: db-config
          mountPath: /tmp
        ports:
        - containerPort: 3000
          name: web
        # env: (any environment variables would go here)
---
apiVersion: v1
kind: Service
metadata:
  name: products-service
spec:
  selector:
    app: products
  type: LoadBalancer
  ports:
  - port: 3000
    targetPort: 3000

Auth Service (Service manifest)

---
apiVersion: v1
kind: Service
metadata:
  name: auth-service
spec:
  selector:
    app: auth
  ports:
  - port: 3000
    targetPort: 3000

Verify cluster nodes Ensure your kubeconfig context is correct and list nodes:

kubectl get nodes

Example cluster output:

NAME                             STATUS   ROLES   AGE   VERSION
ip-192-168-24-32.ec2.internal    Ready    <none>  47h   v1.30.4-eks-a737599
ip-192-168-43-108.ec2.internal   Ready    <none>  47h   v1.30.4-eks-a737599

Install the Telepresence client on your local machine Telepresence provides platform-detecting installers. The example below shows the PowerShell steps for Windows (run PowerShell as Administrator):

Run installer steps as an Administrator. The script will install components and adjust network routes on your machine.

PowerShell installer steps:

# 1. Download the latest windows zip containing telepresence.exe and its dependencies (~50 MB)
Invoke-WebRequest https://app.getambassador.io/download/tel2oss/releases/download/v2.20.0/telepresence-windows-amd64.zip -OutFile telepresence.zip

# 2. Unzip; then remove the zip file
Expand-Archive -Path telepresence.zip -DestinationPath telepresenceInstaller/telepresence
Remove-Item 'telepresence.zip'
cd telepresenceInstaller/telepresence

# 3. Run the installer script to install telepresence and its dependencies.
#    By default it installs to C:\telepresence (customizable via -Path).
powershell.exe -ExecutionPolicy bypass -c ". '.\install-telepresence.ps1';"

# 4. Cleanup unzipped installer files
cd ../..
Remove-Item telepresenceInstaller -Recurse -Confirm:$false -Force

# 5. Telepresence is now installed and you can use telepresence commands in PowerShell.

Verify Telepresence is available:

telepresence --help

Install the Traffic Manager into your cluster Telepresence requires a cluster-side component called the Traffic Manager. The Telepresence client includes Helm and can install the Traffic Manager for you: Install with default values:

telepresence helm install

Install with custom values:

telepresence helm install --values values.yaml

The Traffic Manager is installed into its own namespace (commonly ambassador). Check the namespace and pods after installation:

kubectl get namespace
kubectl get pods -n ambassador

Connect Telepresence to your cluster Telepresence uses the active kubeconfig context. Connect:

telepresence connect

Example successful connection output:

Connected to context arn:aws:eks:us-east-1:195275640053:cluster/telepresence, namespace default (https://C22C5458C5800FD626C361045762447E.gr7.us-east-1.eks.amazonaws.com)

Check Telepresence status

telepresence status

Trimmed example status:

OSS User Daemon: Running
  Version        : 2.20.0
  Executable     : /usr/local/bin/telepresence
  Status         : Connected
  Kubernetes server : https://C22C5458C5800FD626C361045762447E.gr7.us-east-1.eks.amazonaws.com
  Kubernetes context: arn:aws:eks:us-east-1:195275640053:cluster/telepresence
  Namespace      : default
  Manager namespace : ambassador
  Intercepts     : 0 total

OSS Root Daemon: Running
  Version        : v2.20.0
  Remote IP      : 127.0.0.1
  Timeout        : 8s
  Subnets        : (2 subnets)
    - 10.100.0.0/16
    - 192.168.0.0/18

OSS Traffic Manager: Connected
  Version        : v2.20.0
  Traffic Agent  : ghcr.io/telepresenceio/tel2:2.20.0

What Telepresence configures on your machine When connected, Telepresence creates routes for the cluster pod and service CIDRs so local processes can reach cluster IPs and service DNS names directly. Verify these routes with:

ip route

You should see routes pointing cluster subnets to the Telepresence tunnel interface (often tel0 on Linux — interface names vary by OS). List services in the cluster

kubectl get svc

Example services output:

NAME                 TYPE           CLUSTER-IP      EXTERNAL-IP                                                                 PORT(S)           AGE
auth-service         ClusterIP      10.100.10.52    <none>                                                                      3000/TCP          9m39s
inventory-service    ClusterIP      10.100.246.69   <none>                                                                      3000/TCP          46m
kubernetes           ClusterIP      10.100.0.1      <none>                                                                      443/TCP           48m
products-service     LoadBalancer   10.100.169.75   a4519082e1ab846e38b03d9760c9e3b9-515600148.us-east-1.elb.amazonaws.com    3000:30782/TCP    2m7s

Because Telepresence makes your laptop appear to be inside the cluster:

ClusterIP services (e.g., auth-service, inventory-service) are reachable directly by service name or cluster IP.
LoadBalancer services (e.g., products-service) remain externally reachable through your cloud provider’s external IP/DNS.

DNS lookups resolve cluster service names as if you were inside the cluster:

nslookup auth-service

Example:

Server: 127.0.0.53
Address: 127.0.0.53#53

Non-authoritative answer:
Name: auth-service
Address: 10.100.10.52

Make HTTP requests directly to cluster services Curl the auth service by service name and port:

curl http://auth-service:3000

Example JSON response:

{"message":"this is the auth service"}

Call the products (entry point) service with query parameters:

curl "http://products-service:3000/?product_ids=1,2,3"

Example aggregated response (inventory counts included):

{
  "data": [
    {"id":1,"name":"iphone 14","price":900,"category":"electronics","onSale":false,"inventoryCount":893},
    {"id":2,"name":"Samsung 40in TV","price":500,"category":"electronics","onSale":true,"inventoryCount":902},
    {"id":3,"name":"Apple MacbookPro","price":2500,"category":"electronics","onSale":false,"inventoryCount":444}
  ]
}

Access individual pods by pod IP Telepresence also routes pod IPs. First get pod details:

kubectl get pods -o wide

Example:

NAME                                READY   STATUS    RESTARTS   AGE     IP              NODE
auth-depl-69bf897bb9-t9hgg          1/1     Running   0          13m     192.168.32.69   ip-192-168-43-108.ec2.internal
inventory-depl-5f8657c56c-1p5r4     1/1     Running   0          50m     192.168.34.172  ip-192-168-43-108.ec2.internal
products-depl-547767bccd-8zcz9      1/1     Running   0          5m54s   192.168.6.186   ip-192-168-24-32.ec2.internal

Curl a pod IP directly:

curl http://192.168.32.69:3000

Example response:

{"message":"this is the auth service"}

Disconnect Telepresence When finished, quit Telepresence to stop local daemons and remove routes:

telepresence quit
telepresence status

After quitting, telepresence status will show the client is no longer connected. Closing notes Telepresence enables using local development tooling (debuggers, profilers, packet capture, etc.) while interacting with cluster services as if your workstation were inside the cluster network. This simplifies troubleshooting DNS, networking, and service interactions without port-forwarding or building debug images. Links and references

Telepresence documentation: https://www.telepresence.io/
Kubernetes documentation: https://kubernetes.io/docs/

Course Introduction

Telepresence For Kubernetes

Demo Telepresence Basics

Watch Video