ArchiveTeam Warrior — donate idle bandwidth to web preservation, one click

Websites die every day — Geocities, forums, news articles, entire social platforms — and ArchiveTeam is the volunteer swarm racing to save them. The Warrior is their turnkey archiving client: deploy it once, and it joins whichever preservation project currently needs workers, grabbing and archiving pages around the clock. This template turns it into a genuine one-click deploy: deploy → set your leaderboard nickname → open the UI. No Docker CLI, no config files, no reading wikis.

What you get in one click:

The official ArchiveTeam warrior image (atdr.meo.ws/archiveteam/warrior-dockerfile), pinned by digest, served through a Railway-managed build from this repo — updatable, ejectable, yours.
Auth enforced by default. A public, unauthenticated warrior lets any stranger on the internet burn your bandwidth and click around your dashboard. This template auto-generates HTTP_USERNAME / HTTP_PASSWORD per deployment — the deploy form never asks, and anonymous access is rejected.
A persistent volume at /home/warrior/projects. Your nickname, project selection, and settings live in config.json on that volume, so they survive redeploys.
Auto-assignment out of the box. SELECTED_PROJECT=auto means ArchiveTeam's tracker points your warrior at whichever project needs help most right now. Prefer a specific project? Change it in the UI in two clicks.

Setup: 3 steps

Deploy — click the button above; the only prompt is DOWNLOADER, your nickname on the ArchiveTeam leaderboard.
Get your credentials — open your new service's Variables tab and copy the auto-generated HTTP_USERNAME / HTTP_PASSWORD.
Open the UI — visit the Railway domain, log in, and watch your warrior pick up a project and start capturing.

⚠️ Before you deploy: this template consumes bandwidth

Archiving is a bandwidth job. While a project is active, your warrior continuously downloads and uploads web data 24/7, and on Railway that traffic is usage-based billing (egress + ingress) — typically several GB per day, sometimes tens of GB on heavy projects. That is the #1 surprise for new deployers, so decide deliberately: leave CONCURRENT_ITEMS=3 (the gentle default), pause the service in the Railway dashboard whenever you're not actively donating, or pick a lighter project in the UI. The CPU/RAM footprint is small; bandwidth is the real cost.

Deploy and Host

About Hosting

Hosting the ArchiveTeam Warrior on Railway provisions a single Warrior service built from this repo's Dockerfile, which wraps the official atdr.meo.ws/archiveteam/warrior-dockerfile image pinned by digest. The warrior's web UI listens on port 8001 and is exposed on your Railway public domain behind HTTP basic auth (HTTP_USERNAME / HTTP_PASSWORD, generated per deployment via Railway's secret()). A Railway volume is mounted at /home/warrior/projects so the warrior's config.json — nickname, selected project, and UI settings — persists across redeploys. First-boot behavior is seeded from image defaults (SELECTED_PROJECT=auto, CONCURRENT_ITEMS=3, SHARED_RSYNC_THREADS=20); afterwards the warrior reads its configuration from config.json on the volume, which is exactly what makes your settings survive restarts.

Why Deploy

Deploying the warrior manually means running Docker commands on a machine you keep awake, remembering to add auth, and losing your configuration every time the container is replaced. On Railway you get a one-click deploy with auth already enforced, a volume that keeps your identity and project selection across redeploys, automatic restarts (restartPolicyType: ALWAYS), a public URL with TLS, and an observable dashboard with logs and usage metrics. You can pause the warrior whenever you don't want it working — something a home server doesn't make as easy.

Common Use Cases

Digital preservation volunteering — join the ArchiveTeam swarm and help archive at-risk websites, tracked on the public leaderboard under your nickname.
Self-hosted archiving rig — run a warrior in the cloud instead of a spare laptop or Raspberry Pi at home, with restart-on-failure and remote access via the auth-gated web UI.
Team or community contribution — a shared, credentialed warrior instance for a community that wants its nickname on the leaderboard without everyone installing Docker.
On-and-off donating — pause the service during bandwidth-costly periods and resume with all settings intact thanks to the persistent volume.

Dependencies for

The template has zero external service dependencies — the warrior talks directly to ArchiveTeam's tracker and rsync servers from the public internet, and all state lives in its own Railway volume.

Deployment Dependencies

Warrior service — built from this repo (Dockerfile pinning atdr.meo.ws/archiveteam/warrior-dockerfile by digest). Public domain on port 8001. Variables: DOWNLOADER (your leaderboard nickname, asked at deploy time), HTTP_USERNAME and HTTP_PASSWORD (auto-generated per deployment — view them in the Variables tab).
Volume — mounted at /home/warrior/projects, stores projects/config.json (nickname, project selection, UI credentials) so settings survive redeploys.

⚠️ Running cost note: the warrior's job is moving web data. Expect usage-based bandwidth charges (egress + ingress) while a project is active — see the cost warning above for how to keep them under control.