Practical Lab

This lab guide provides step-by-step instructions for implementing four key hybrid cloud scenarios. These exercises are designed to give you hands-on experience with networking, application deployment, disaster recovery, and multi-cloud Kubernetes.

Scenario 1: Networking - The Universal Overlay Network

Goal: Create a secure, encrypted overlay network (Mesh VPN) that connects resources across Local, Linode, and GCP using Tailscale or Wireguard.

Step 1: Install Tailscale on All Nodes

Local Machine: Download and install Tailscale from tailscale.com.

Linode VPS:

curl -fsSL https://tailscale.com/install.sh | sh
sudo tailscale up

GCP Instance:

curl -fsSL https://tailscale.com/install.sh | sh
sudo tailscale up

Step 2: Authenticate and Verify

Follow the authentication links provided in the terminal for each node.

Once authenticated, verify connectivity from your local machine:

tailscale status
ping <linode-tailscale-ip>
ping <gcp-tailscale-ip>

Step 3: Enable Subnet Routing (Optional)

If you want to access the entire VPC/Local network behind a node:

On the node: sudo tailscale up --advertise-routes=10.0.0.0/24 (replace with your subnet).
In the Tailscale Admin Console, enable the "Subnet Route" for that device.

Scenario 2: Application Deployment - Distributed Tiered Architecture

Goal: Deploy a 3-tier application where components are distributed across different environments.

Database: PostgreSQL (Local)
App Server: Flask/Node.js (Linode)
Object Storage: Cloud Storage (GCP)

Step 1: Set up Local Database

Install and start PostgreSQL.
Create a database and a user that allows connections from the Tailscale IP of the Linode server.
Update pg_hba.conf and postgresql.conf to allow remote connections on the Tailscale interface.

Step 2: Set up GCP Cloud Storage

Create a Bucket in GCP.
Create a Service Account with Storage Object Admin permissions.
Download the JSON key file.

Step 3: Deploy Application on Linode

Clone your application code.
Configure the application to use:
The Local Machine's Tailscale IP for the database connection.
The GCP Service Account key for storage operations.
Start the application and verify it can read/write to both the local DB and GCP storage.

Scenario 3: Disaster Recovery & Backup - Tiered 3-2-1 Strategy

Goal: Implement a robust backup strategy where data flows from Local to Linode to GCP.

Local: Primary Data
Linode: Hot/Warm Backup (S3-compatible Object Storage)
GCP: Archive (Cloud Storage Archive Class)

Step 1: Local to Linode (Hot Backup)

Use rclone to sync local data to Linode Object Storage.

rclone sync /path/to/local/data linode-s3:my-backup-bucket

Step 2: Linode to GCP (Archive)

Use rclone or a scheduled script to move older backups from Linode to GCP Archive.

rclone copy linode-s3:my-backup-bucket gcp-storage:my-archive-bucket --storage-class ARCHIVE

Step 3: Verify Restoration

Delete a local file.
Restore it from the Linode backup.
Simulate a total Linode failure and restore from the GCP archive.

Scenario 4: Multi-cloud Kubernetes - The Spanning K3s Cluster

Goal: Create a single Kubernetes cluster with nodes in different clouds.

Master Node: Linode
Worker Node 1: Local Machine
Worker Node 2: GCP Instance

Step 1: Initialize K3s Master (Linode)

Install K3s with the Tailscale IP as the advertise address:

curl -sfL https://get.k3s.io | INSTALL_K3S_EXEC="--node-ip=<linode-tailscale-ip> --flannel-iface=tailscale0" sh -

Retrieve the node token: sudo cat /var/lib/rancher/k3s/server/node-token

Step 2: Join Worker Nodes (Local & GCP)

Install K3s agent on workers, pointing to the Linode master:

curl -sfL https://get.k3s.io | K3S_URL=https://<linode-tailscale-ip>:6443 \
K3S_TOKEN=<node-token> INSTALL_K3S_EXEC="--node-ip=<worker-tailscale-ip> --flannel-iface=tailscale0" sh -

Step 3: Verify Cluster Status

On the Master node:
```
kubectl get nodes -o wide
```
You should see all three nodes (Linode, Local, GCP) in a Ready state, communicating over the overlay network.