Upgrade
Back up configuration, review release notes, pin the new release tag, rebuild with fresh base images, verify readiness, and test representative URLs before cutover.
Open-source engine
Own the renderer.
Own the operations.
Prerender Buddy Engine is an Apache-2.0 Docker service for teams that want crawler rendering inside infrastructure they operate.
Quick start
A small, auditable core.
Version 0.1 runs one Chromium process with bounded concurrency, in-memory caching, duplicate-render coordination, URL controls, and no external telemetry.
Docker Composelocalhost:3000
git clone https://github.com/kopachlager/prerenderbuddy-engine.git
cd prerenderbuddy-engine
cp .env.example .env
# Set a 32+ character token and exact allowed domains in .env
docker compose up --build -d
curl http://127.0.0.1:3000/readyConfiguration
Secure defaults first.
PRERENDER_TOKENRequired server-side secret with at least 32 characters.
ALLOWED_DOMAINSRequired exact hostnames by default; list apex and www separately.
RENDER_MAX_CONCURRENCYBound parallel browser work to the capacity you have measured.
RENDER_TIMEOUT_MSBound page execution instead of allowing long-lived renders.
CACHE_*Size the process-local cache by entries, bytes, and TTL.
ALLOWED_ORIGINSLeave empty unless browser CORS access is deliberately required.
Production readiness
The container is one part of the service.
- 01Keep the engine private behind TLS or a trusted server-side integration.
- 02Use exact ALLOWED_DOMAINS values and block access to private networks at the infrastructure layer.
- 03Start with 2 vCPU, 2 GB RAM, and render concurrency between 2 and 4; measure representative pages.
- 04Monitor health, readiness, capacity, memory, latency, failures, and restarts before sending crawler traffic.
- 05Pin a release tag, test updates in staging, and keep a documented rollback path.
Support boundary
Self-hosted core or managed outcome.
The open engine is community-supported software. The managed product includes operated services that are not part of the repository.
| Capability or responsibility | Open engine | Managed service |
|---|---|---|
| Rendering engine and Docker image | ||
| Crawler routing and production integration | ||
| TLS, network isolation, firewall, and secrets | ||
| Monitoring, alerting, scaling, and incidents | ||
| Managed DNS, dashboard, billing, and support |
For the open engine, your team owns deployment, crawler routing, uptime, scaling, monitoring, backups, upgrades, security response, and incident handling. No service-level guarantee is included.
Troubleshoot
Use container logs for startup failures, /ready for Chromium state, HTTP status codes for token, domain, and capacity errors, and metrics for memory and concurrency pressure.
Harden
Keep the token server-side, isolate outbound networking, deny metadata and private ranges, rate-limit the proxy, and separate unrelated tenants with different containers and tokens.
Choose the ownership model