Elm static‑HTML output combined with Node.js port‑based SSR: evaluating cost trade‑offs for low‑traffic workloads

Answer to the core question

For a typical low‑traffic Elm site, the break‑even point lies around 300 k requests per month (≈10 k requests per day). Below that threshold, the cost of a static host (≈$0.02/GB outbound) is lower than the monthly cost of a minimal Node.js instance (≈$36/month if running 24/7). Above that volume, the Node.js server can become cost‑effective, especially if you can shut it down during idle periods or use a serverless function.

CPU impact of Elm ports

Each port call serializes data via JSON.stringify/parse, adding roughly 0.5 ms of CPU time per 10 KB of payload on a modern CPU. A 200 KB payload therefore costs ~10 ms per request. If ports are invoked on every request, 10 k requests/day would consume ~100 s of CPU time per day – usually negligible on a low‑cost instance, but it scales linearly with payload size and frequency.

Limitations of elm‑static‑html

• Experimental tool, currently only supports Elm 0.19.1.
• No code‑splitting, dynamic imports, or hot‑reloading.
• Breaking changes in Elm 0.19.2 could force a rebuild and migration of the static‑HTML generator.

Operational overhead

A Node.js process requires patching, uptime monitoring, and potential scaling. Static hosts (Netlify, Vercel, CloudFront) provide zero‑maintenance, auto‑scaling, and built‑in CDN caching, drastically reducing operational complexity.

Recommendation for your scenario

If your site expects under 10 k requests/month and port payloads stay <200 KB, static hosting with elm‑static‑html is the simplest and cheapest path.

If you anticipate more than 300 k requests/month or need to deliver larger dynamic data via ports, a lightweight Node.js server (or serverless function) becomes cost‑effective, provided you can shut it down during low‑traffic periods.

Next steps

1. Measure your actual outbound traffic and request volume on a static host (e.g., CloudFront or Netlify analytics).
2. Deploy a minimal Node.js server with an Elm port, instrument CPU usage (Node's --cpu-prof or a profiler), and log the average port payload size.
3. Plug real numbers into the cost formulas above, adjusting for your provider’s pricing and CDN caching.
4. Test elm‑static‑html with the current Elm compiler to confirm compatibility and identify any build errors.

Missing diagnostic detail: What is the average size of the data you plan to exchange through Elm ports on each request? Knowing this will refine the CPU cost estimate and may shift the break‑even point.