How I replaced Google Photos, Google Drive, and Google Home with a self-hosted Raspberry Pi 5 setup - and what the full technical stack looks like.

There's a quiet moment every tech-literate person eventually hits. You open your cloud storage dashboard, see the number creeping up, and think: I'm paying a subscription fee, every month, forever, just to store my own photos and files on someone else's computer.

That was me, but with a bit of added weight. It's not just my data I'm responsible for.

Over the years, I've quietly become the unofficial digital curator for my entire family. The person everyone sends photos to after a birthday party. The one who backs up the wedding videos. The one who scans and stores the important documents, passports, contracts, and sentimental things so they don't get lost. My parents, siblings, extended family: if something matters and it's digital, there's a good chance it eventually lands with me.

That's a responsibility I take seriously. And for a long time, Google was my answer - until the storage bill started quietly creeping past 2TB, and I started thinking more carefully about what it actually means to hand all of that data over to "big tech".

So I built my own home server. A tiny, almost silent box sitting in my house that now handles everything Google Drive and Google Photos did, plus more, for a one-time hardware cost of £340.

This is the story of how I did it, why I did it, and exactly how you can too.

Why I Did It

The cost was the trigger, but it wasn't the only reason.

When you use Google Photos, Google Drive, or any cloud storage service, your files live on their servers. You're trusting a corporation to keep them safe, not snoop through them, not change their pricing, and not shut down the product one day. Google has a long history of killing beloved products. Google Photos itself famously ended its unlimited free tier in 2021. As well as this, recent AI trends and auto opt-in data processing had me thinking there had to be a better way for my own photos and videos to remain private to me.

But the thing that really focused my mind was thinking about what I was actually storing. This isn't just my holiday photos. It's:

• Videos and pictures my family have taken of my childhood
• Scanned copies of family documents that only exist in one place
• Every photo & video from the last family wedding
• Memories that, if lost, are probably gone forever

Being the person everyone trusts with that data means I can't afford a casual approach to storage. If Google decides to raise prices again, change a policy, or simply shut down a product, I don't have a fallback. That felt like an unacceptable risk once I really sat down and thought about it.

I wanted:

• My data on my hardware in my house
• No monthly fees beyond sips of electricity
• Access from anywhere not just on my home network
• Genuine resilience multiple backups, multiple locations, no single point of failure

The good news? All of this is completely achievable for a normal person with a bit of patience. You don't need to be a software engineer like me to get it up and running.

What It Looks Like Day To Day

Before getting into the technical weeds, here's what I actually use this thing for:

Immich replaces Google Photos. It looks almost identical, automatic photo backups from my phone, face recognition, map view, albums, the works. The app is free, open source, and genuinely excellent.

Nextcloud replaces Google Drive. Files, documents, calendar sync, contacts, all running on my server, syncing to my devices exactly like Google Drive does.

Home Assistant handles home automation. Lights, sensors, routines. Think of it as a private, local version of Google Home that doesn't need to connect to some cloud server far away to simply turn on my lights.

And the best part? I can access all of it from my phone, from a cafe, from anywhere in the world, through my own domain.

The Hardware

The brain of this setup is a Raspberry Pi 5, a small single-board computer roughly the size of a deck of cards. It sips electricity, runs silently, and is powerful enough to handle everything I need.

Here's the full hardware list and what I paid in December 2025:

⚠️ Heads up: Prices have risen since purchase. AI demand has pushed up the cost of compute and memory. Budget a little more if buying today.

The Software Stack

This is where people get intimidated. But the software is actually the easy part; most of it is one-click installs.

CasaOS - Your Dashboard

CasaOS is a slick web interface that sits on top of the Pi and lets you install and manage apps like Immich, Nextcloud, and Home Assistant with a single click. Think of it as an app store for your server. You access it by typing your Pi's local IP address into a browser.

Cloudflare - Your Front Door to the World

This is the clever bit that lets you access your server from anywhere, securely, without opening up your home network to the internet.

Cloudflare Tunnels creates an encrypted tunnel between your Pi and Cloudflare's network. When you visit yourdomain.com, Cloudflare securely routes that request to your Pi without your home IP address ever being exposed. It's free, and it's genuinely brilliant.

Nginx Proxy Manager - The Traffic Director

When a request comes in for yourdomain.com, something needs to direct it to the right app running on your Pi. Nginx Proxy Manager does exactly that. It's a visual interface for managing which web address goes to which app.

The Apps

What It Costs to Run

Electricity: A Raspberry Pi 5 under normal load uses roughly 5-10W. At UK energy prices, that's pennies per day - around £3-5/month at most.

Domain: My domain costs a few pounds a year through Cloudflare.

Cloudflare Tunnels: Free.

Backups (planned todo): AWS S3 Glacier Deep Archive for long-term cold storage costs roughly $0.00099/GB/month - that's about £1.50/month for 2TB.

So the ongoing cost is roughly £5-7/month - compared to Google One's 2TB plan at £8.99/month, and rising since I've almost filled my 2TB (next tier is the £18.99 a month 5TB which is why I really wanted to get this done before I hit the cap!).

Is It Worth It?

Yes - with one honest caveat.

There is a setup cost in time. Getting everything running took several evenings of reading, tinkering, and occasionally breaking things and starting over. It's not a 20-minute job.

But the context matters here. When you're just managing your own photos and files, maybe the risk of losing them to a cloud outage or a price hike is something you can accept. When you're the person holding irreplaceable family memories - wedding videos, childhood photos, documents that exist in one form only - that sways the decision. I could keep paying more and more in cloud storage costs, or I could try to take things into my own hands and build my own cloud storage.

Once it's running, it just works. And there's a real peace of mind in knowing that the photo from your childhood, or the scan of your parents' old documents, isn't one cancelled subscription away from disappearing.

Your data is yours. It should live somewhere you control.

Full disclosure, for redundancy, I'm still paying for Google storage until I get myself some more storage in the form of HDD drives & set up the S3 Glacier back-ups.

🔧 Technical Deep Dive

Everything below is for the people who want the full picture - every config, command, and where I was getting stuck.

Stack Overview

Raspberry Pi OS Lite (Bookworm)
└── CasaOS (dashboard + app management)
    ├── Immich
    ├── Nextcloud
    ├── Home Assistant
    └── Portainer
cloudflared (bare metal systemd service)
Nginx Proxy Manager (bare metal or container)

Networking:

Internet → Cloudflare DNS → Cloudflare Tunnel → cloudflared on Pi → Nginx Proxy Manager → App container

Operating System

Raspberry Pi OS Lite (Bookworm) - headless, no desktop environment. All access via SSH & then exposing ports for various software UI's.

# Install a useful system monitor for SSH sessions
sudo apt install bpytop

Booting from NVMe (Not SD Card)

SD cards are slow and will fail under sustained write load. Boot from the NVMe SSD instead.

Tool: rpi-clone by Jeff Geerling

# Step 1: Reflash SD card and boot normally

# Step 2: Format the NVMe drive cleanly
sudo nvme format /dev/nvme0n1 --force

# Step 3: Verify settings
sudo raspi-config

# Step 4: Clone SD → NVMe (rpi-clone, installed from git)

# Step 5: Shut down, remove SD card, confirm Pi boots from NVMe alone

💡 Note: rpi-clone handles the bootloader config automatically. Don't try to do this manually with dd unless you know what you're doing.

CasaOS

Installed on the bare OS. Manages Docker containers via a web UI.

Local access:

• http://192.168.1.183:88 - ethernet (eth0)
• http://192.168.1.184:88 - Wi-Fi (wlan0)

Fix: Port Conflict with Nginx on Reinstall

Symptom: After a reinstall, you see Nginx's default "Congratulations" page instead of the CasaOS dashboard.

Cause: A previous Nginx install is still running on port 80, blocking CasaOS. This threw me off, and I ended up reflashing and installing everything a couple of times here until I figured out the issue.

Fix:

sudo vi /etc/casaos/casaos.conf
# Add under [server]:
# HttpPort = 88

systemctl restart casaos
systemctl status casaos

# Verify at http://<PI_IP>:88

Cloudflare Configuration

DNS

All DNS for mydomain.com is managed in the Cloudflare dashboard at dash.cloudflare.com.

SSL

SSL certificates are issued via Let's Encrypt using a Cloudflare API token, handled through Nginx Proxy Manager.

Cloudflare Tunnels

Tunnels expose internal services to the internet without opening ports on your router or exposing your home IP.

Dashboard: one.dash.cloudflare.com

Setup rules:

• The tunnel name must be unique and must not conflict with any existing *.mydomain.com subdomain
• Each published application route automatically creates a CNAME DNS record in Cloudflare

cloudflared on the Pi

cloudflared runs as a bare-metal systemd service (not in a container — keeps it resilient).

# Authenticate and pull down the origin certificate
cloudflared tunnel login

# Install as a systemd service
sudo cloudflared service install <TUNNEL_ID>

Nginx Proxy Manager

Local access: http://192.168.1.183:81

Acts as a reverse proxy — routes incoming requests by hostname to the correct internal container port.

For each Cloudflare Tunnel route, add a matching Proxy Host in Nginx Proxy Manager:

SSL termination is handled here via the Cloudflare API token.

Note: you can have whatever subdomain of your choosing as long as it's unique. I've used drive.mydomain and photos.mydomain as my muscle memory from Google's offerings is strong and it's also what I'm replacing with my self-hosted options.

Nextcloud

Trusted Domains

Nextcloud rejects requests from any domain not in its trusted list. Add yours to the config file.

File: DATA/AppData/big-bear-nextcloud/html/config/config.php

'trusted_domains' =>
array (
  0 => 'localhost',
  1 => '192.168.1.184',
  2 => 'drive.mydomain.com',
),

Add any additional IPs or subdomains you want to access Nextcloud from.

Portainer

Used alongside CasaOS for lower-level container management — useful for inspecting logs, environment variables, network configs, and volume mounts that CasaOS abstracts away.

Backup Strategy — Why This Is Probably The Most Important

I want to spend a bit more time on this than a typical homelab post would, because it's the part most people underestimate, as I did initially & the part where I refuse to cut corners.

I'm not just storing my own data. I'm the person my family trusts with photos from birthdays, weddings, and events that can never be re-shot. Important documents. Irreplaceable things. Losing that data because I only had one copy on one device isn't an option here & why I've dug into what "good" looks like in this area.

The 3-2-1 Rule

The industry standard for data resilience is the 3-2-1 Rule [Backblaze, Druva]:

• 3 total copies of your data (your primary + two backups)
• 2 different types of storage media (e.g. NVMe SSD + external HDD)
• 1 offsite copy (stored remotely or in the cloud)

The reason for the rule is simple: any single backup can fail. A fire, a flood, a hardware failure, a ransomware attack — any of these can wipe out everything stored in one place. The 3-2-1 rule ensures no single event can take everything.

My Hybrid Backup Strategy (In Progress)

I'm implementing a layered approach that satisfies the 3-2-1 rule while keeping ongoing costs low:

Layer 1 — Primary (NVMe SSD): The live server. All data lives here, all apps read from here.

Layer 2 — Local HDD Enclosure: A separate external hard drive connected to the Pi. Scheduled snapshots copy the NVMe data here on a regular cadence. If the SSD fails, everything is recoverable within minutes from the same device. This is the next hardware addition I'm working on.

Layer 3 — AWS S3 Glacier Deep Archive: Weekly or fortnightly syncs push data offsite. This is the safety net for worst-case scenarios — fire, theft, or catastrophic hardware failure.

Glacier Deep Archive is designed for data you rarely need to touch but absolutely cannot lose. At roughly £1.50/month for 2TB, it's the cheapest insurance policy I'll have ever bought.

💡 Note: I'm still exploring additional backup options and tools — this is an evolving strategy rather than a finished one.

What's Next

• HDD enclosure for local backups (Backup Layer 2)
• Automated snapshot scheduling (NVMe → HDD)
• AWS S3 Glacier Deep Archive sync (Backup Layer 3)
• Explore additional hybrid backup tools and strategies
• Run LLM agents on the server
• Explore phone-triggered automations

Got questions, or are you also the unofficial family data hoarder? Drop a comment — or message me on LinkedIn.