Build Your Own AWS at Home

Recreate cloud services using Raspberry Pi and old desktop computers. High-availability, open-source, and completely under your control.

user@homecloud:~
Type 'help' to see available commands
homecloud help
EKS: Running (K3S v1.32.3)
S3: Running (MinIO v5.4.0)
homecloud deploy rds
Deploying RDS service with CloudNativePG...
Creating PostgreSQL cluster 'homecloud-db'
✔ Cluster created successfully

Planned Services

Recreating AWS services one by one using open-source alternatives that you can run at home.

Amazon S3

-> MinIO Object Storage

Object storage service implemented using MinIO, providing S3-compatible API for your applications.

MinIO S3-compatible Multi-node

Amazon EKS

-> K3S

Managed Kubernetes service using K3S, a lightweight Kubernetes distribution perfect for edge and home use.

K3S Lightweight ARM64 support

AWS RDS

-> CloudNativePG

Managed PostgreSQL service using CloudNativePG, with automated backups, failover, and scaling.

PostgreSQL HA Backups

ElasticCache Redis

-> Bitnami Redis

Redis-based caching service using Bitnami's Redis Helm chart, with clustering and persistence options.

Redis Bitnami Cluster

Secrets Manager

-> Binami Sealed Secrets

Secure secret management using Sealed Secrets, encrypting your secrets for safe storage in Git.

Sealed Secrets GitOps Encryption

Container Registry (ECR)

-> Zot OCI Registry

Private container registry using Zot, an OCI-compliant registry with support for image signing and vulnerability scanning.

Zot OCI Private Registry

Designed for Consumer Use Hardware

Optimized to run on affordable, energy-efficient hardware like Raspberry Pi and repurposed desktop computers.

Raspberry Pi

ARM64 support with optimized images for Pi 4 and newer models.

x86 Desktops

Leverage old hardware with multi-node clustering for HA.

Low Power

Designed to run 24/7 without excessive energy costs.

Compact

Entire stack can run on a single device or scale horizontally.

Hardware Requirements

Minimum

8GB RAM, 4-core CPU, 64GB storage (Raspberry Pi 4B) * 1

Recommended

8GB RAM, 4-core CPU, 64GB+ storage (Raspberry Pi 4B) * 3

Any x86 Desktop (With Nvidia Graphic Cards) * 1

High Availability Storage

3+ nodes with HDD or SSD (Longhorn)

Power Consumption

~5W per Raspberry Pi, ~100W per x86 node (If GPU Exist)