Turn an Old PC Into a Linux Home Server With Home Assistant, Jellyfin, Kopia and Paperless-ngx

Before you consign that aging silicon to the e-waste heap, consider why it feels slow in the first place. A computer that chokes on a bloated desktop environment and three dozen active browser tabs often has plenty of muscle left when stripped of interactive duties. As Yadullah Abidi recently highlighted in MakeUseOf, installing a lightweight Linux distribution, setting up Docker or virtual machines, and handing an older PC a dedicated background role can turn a sluggish daily driver into a remarkably capable home server. Specifically, four free, open-source applications—Home Assistant, Kopia, Paperless-ngx, and Jellyfin—offer compelling reasons to keep an old box plugged in and humming on your local network.

Of course, any veteran systems administrator will tell you that "free software on old hardware" comes with architectural fine print. Before you start pulling dusty towers out of the closet, let us break down how these four tools work, what hardware and operating system constraints you will actually face, and how to deploy them without creating a brand-new IT headache for yourself.

Home Assistant: The Local Smart Home Commander​

If your smart home currently relies on a fragile truce between four different smartphone apps and three proprietary cloud hubs, you are essentially renting control of your own light switches. MakeUseOf points to Home Assistant as a premier upgrade for unifying cross-brand smart lights, plugs, thermostats, cameras, and sensors into a single interface capable of running custom automations—like triggering hallway lighting on motion after sunset, turning LED strips into visual alarms, or shutting down appliances when the house is empty.

The Real-World Architecture of "Local Control"​

The headline appeal of Home Assistant is local execution. According to official Home Assistant documentation, devices communicating over local protocols—specifically Zigbee, Z-Wave, Matter, Thread, and ESPHome—continue functioning and running automations even when your internet connection drops or a manufacturer's cloud platform goes dark.

However, local control is not magic, and it is not automatic across every gadget you own:

• Cloud-tethered devices still need the internet: Official Home Assistant documentation notes that if a device only speaks to its manufacturer's cloud servers, Home Assistant cannot reach it during an internet outage. Similarly, Home Assistant Cloud features like remote out-of-home access and integrations with Amazon Alexa, Google Home, or Apple Home require an active internet connection.
• Radio dongles and border routers: Your old PC does not speak Zigbee, Z-Wave, or Thread out of the box; as MakeUseOf notes, those protocols require plugging a compatible USB radio dongle into the PC. Furthermore, Home Assistant's Matter documentation clarifies that Matter itself is an IPv6 application-layer control protocol running over Wi-Fi, Ethernet, or Thread rather than a standalone radio technology. Using Thread-based Matter devices requires a Thread border router on the network, and commissioning new Matter devices generally uses Bluetooth via the Home Assistant Companion app on your phone rather than the PC's internal Bluetooth adapter.

Choosing Your Installation Path​

How you install Home Assistant dictates what else that old PC can do. MakeUseOf notes that an aging x86 or x64 machine can easily run Home Assistant inside a Docker container or virtual machine, or you can flash the dedicated Home Assistant Operating System (HAOS) image on compatible 64-bit UEFI PCs with Secure Boot disabled.

Here is the critical distinction from the official Generic x86-64 installation guide:

1. Dedicated HAOS Image: This turns the PC into an exclusive appliance. It wipes the drive, requires a 64-bit UEFI system with Secure Boot disabled, and demands a boot medium using 512-byte logical sectors (512n or 512e), meaning native 4K sector (4Kn) NVMe or SATA drives will fail to boot. It also requires an active Ethernet internet connection on first boot to download Home Assistant Core.
2. Home Assistant Container (Docker): If you want your old PC to run Kopia, Paperless-ngx, and Jellyfin alongside Home Assistant on a single Linux host, you will likely run Home Assistant Container via Docker Engine (version 23.0.0 or later; Docker Desktop is not supported). Be aware that the standalone container installation does not include the Home Assistant Supervisor or access to the built-in add-on ("apps") store, and you must manually map USB radio devices (such as /dev/tty*) into the container and allow an extended shutdown timeout so its database closes cleanly.

Section Summary: Home Assistant can liberate your smart home from vendor clouds, provided you pair it with locally communicating devices, add the necessary USB radios, and choose between a dedicated HAOS appliance or a shared Docker host.

Kopia: Encrypted, Deduplicated Household Backups​

Why leave backup tasks to manual external hard drive rituals that nobody in the house remembers to perform? MakeUseOf highlights Kopia as a way to transform an always-on PC into a central backup hub for the entire household.

How Kopia Handles Your Data​

Available for Windows, macOS, and Linux with both graphical (GUI) and command-line (CLI) interfaces, Kopia creates point-in-time snapshots of your critical files and directories rather than monolithic full-disk images. According to Kopia's official documentation and MakeUseOf's overview, its core strengths include:

• Incremental, deduplicated storage: After the initial backup, Kopia only stores new or modified data, and its content deduplication prevents storing duplicate data blocks shared across multiple files or even multiple client computers backing up to the same repository.
• End-to-end encryption by default: Repositories are encrypted using a user-chosen password that remains on your machine, paired with configurable compression, snapshot scheduling, and retention policies.
• Flexible storage targets: You can save snapshots to local disks, network-attached storage (NAS), a dedicated Kopia Repository Server on your old PC, SFTP or WebDAV servers, or cloud storage providers including Amazon S3-compatible buckets, Azure Blob Storage, Backblaze B2, and Google Cloud Storage (as well as experimental Rclone integration for services like OneDrive, Google Drive, and Dropbox).

The Golden Rule of Backup Architecture​

Setting up Kopia on your primary Windows desktops and laptops to push scheduled snapshots to a high-capacity drive on your old PC is a great start. However, MakeUseOf rightly warns that this local server must never be your only backup.

From an IT resilience standpoint, an always-on PC sitting in the same building as your primary computers is vulnerable to the exact same power surges, fires, thefts, or ransomware outbreaks hitting your network. A sound backup strategy follows the classic 3-2-1 rule: keep multiple copies across at least two storage mediums, and configure Kopia to replicate or send a secondary encrypted repository to offsite cloud storage. Also, guard your Kopia repository password carefully—end-to-end encryption means there is no "forgot password" link if disaster strikes.

Section Summary: Kopia turns an old PC with a large hard drive into an automated, deduplicated backup destination for every Windows, Mac, and Linux machine in your home, so long as you pair it with an offsite backup copy.

Paperless-ngx: Taming the Paper Mountain​

Every household has a "drawer of doom" stuffed with crumbling thermal receipts, appliance manuals, warranty cards, and tax statements. Finding a single serial number in that pile takes twenty minutes; finding it in a poorly named folder of PDF scans takes nineteen. MakeUseOf recommends Paperless-ngx to turn an old PC into a private, searchable document archive.

How OCR and Auto-Classification Work​

When you feed scanned documents into Paperless-ngx, it runs Optical Character Recognition (OCR) so you can search for any company name, address, product model, or phrase buried inside the page text rather than relying on file names. It also organizes files using tags, correspondents, document types, storage paths, and custom fields, and uses an automatic matching algorithm to learn from your existing archive and apply metadata to newly imported scans.

According to official Paperless-ngx documentation, the software runs on Linux (Windows is not supported directly as a bare-metal host) and is most easily deployed via its official Docker Compose setup, which automatically manages dependencies like the Redis/Valkey message broker and OCRmyPDF/Tesseract engines.

Tuning Paperless-ngx for Aging Processors​

MakeUseOf points out a crucial hardware truth: OCR briefly taxes an older CPU when a document is ingested, whereas browsing and searching the indexed archive afterward is light work.

If you are deploying Paperless-ngx on a particularly weak CPU or a machine with limited RAM, official Paperless-ngx setup and configuration documentation offers several practical tuning levers:

• Skip redundant OCR: Configure OCR to skip text recognition (--skip-text) on digital PDFs that already contain embedded text, reserving heavy CPU cycles for actual image scans.
• Throttle background workers: Lower the worker and thread counts (for instance, 2 workers and 1 thread) so a batch of scanned receipts does not starve other Docker containers on the server.
• Use SQLite on low-spec hardware: Sticking with the default SQLite database rather than running a full PostgreSQL or MariaDB container saves system memory.
• Mind your account security and file shares: Because the initial Paperless-ngx superuser account has unrestricted access to all documents, the official guide recommends creating a standard user account for daily browsing. Additionally, if your consumption folder sits on an NFS or SMB network share where native Linux inotify file notifications fail, set PAPERLESS_CONSUMER_POLLING_INTERVAL so Paperless checks for new scans reliably.

Section Summary: Paperless-ngx replaces paper clutter and cloud document services with a local, full-text searchable archive that runs smoothly on older Linux/Docker servers once initial OCR ingestion finishes.

Jellyfin: Cutting the Cord on Media Server Subscriptions​

Rounding out the quartet is Jellyfin, an open-source media server that organizes your personal library of movies, TV shows, and music and streams them to phones, PCs, and smart TVs without putting core features behind a subscription paywall.

Direct Play vs. The Transcoding Bottleneck​

Streaming a video file across your local network takes minimal processing power when the client device natively supports the file's video codec, audio format, and subtitles—a scenario known as Direct Play. The trouble starts when a TV, phone, or web browser cannot play the format directly, forcing the server to transcode (convert) the video on the fly.

Software transcoding on an ancient dual-core or quad-core CPU will quickly bring the machine to its knees. As MakeUseOf notes, the saving grace for many older PCs is Intel Quick Sync Video (QSV), a dedicated hardware media block built into many Intel integrated GPUs that can accelerate video decoding and encoding without overwhelming the main CPU cores.

Verifying What Your Old Intel Chip Can Actually Do​

Before assuming your vintage PC will effortlessly transcode 4K video, check Jellyfin's official hardware acceleration documentation against your exact processor generation:

Furthermore, Jellyfin's documentation stresses that you must use the custom jellyfin-ffmpeg binary (which comes bundled with official Jellyfin Docker images, Debian packages, and Windows installers) to achieve full hardware acceleration. Even then, tasks like burning in complex subtitles can still fall back to partial acceleration and tax your CPU.

Section Summary: Jellyfin turns an old PC into a subscription-free Netflix for your personal media, performing best when you favor Direct Play clients or pair a 5th-gen-or-newer Intel chip (or VA-API setup) with formats your integrated GPU natively supports.

Practical Deployment Checklist for Windows Forum Readers​

If you are ready to rescue an old desktop or laptop from retirement, resist the urge to install all four services on day one. Follow MakeUseOf's advice: pick one application that solves an immediate frustration, get it stable, and expand from there.

1. Choose your host architecture first: Decide whether this machine will be a single-purpose Home Assistant OS appliance (which claims the entire PC) or a general-purpose Linux server running Docker Engine and Docker Compose (which lets you run Home Assistant Container, Kopia, Paperless-ngx, and Jellyfin side by side).
2. Hardwire the network: Put the server on Gigabit Ethernet rather than Wi-Fi, especially if it will ingest Kopia backups or stream Jellyfin video.
3. Audit your storage and boot settings: Ensure your BIOS is set to UEFI mode (and disable Secure Boot if installing HAOS), check drive health before trusting it with Paperless-ngx archives or Kopia repositories, and use an SSD for the OS and Jellyfin transcode cache.
4. Back up the server itself: Whether it is your Home Assistant automation configurations, your Paperless-ngx data and media volumes, or your Kopia repositories, make sure your new home server is part of a multi-copy backup strategy.