Infrastructure, Funding, and Temporary Downtime

A transparent status update on limited hosting budget, cold starts, and why some EDHcheck services may be temporarily unavailable after idle periods.

A quick transparency update about how EDHcheck is currently running behind the scenes.

The project is still operating on a limited budget. To keep it online and improving, parts of the stack use cost-efficient infrastructure that can scale down during inactivity. That helps keep monthly costs manageable, but it can introduce occasional startup delays or short windows where a feature appears unavailable.

What this means in practice

If the site has been idle for a while, you may notice one or more of the following:

  • First request delay: the first action (analyze, comments, profile update) can take longer than usual.
  • Temporary "unavailable" responses: some endpoints may return an error while services wake up.
  • Intermittent import/API failures: third-party dependency hiccups can feel worse during warm-up periods.
  • Retry success pattern: a second attempt after a short wait often works normally.

Things to keep in mind as a user

  • If something fails once, wait 10 to 30 seconds and retry.
  • For imports, keep a text-export fallback ready when possible.
  • If comments or profile actions fail once, refresh and try again.
  • Report persistent failures in comments so I can prioritize fixes.

Why this setup exists right now

The current goal is simple: keep EDHcheck live, useful, and improving without overcommitting infrastructure costs too early. This trade-off lets development continue while traffic and support grow.

As the project gets more support, the plan is to increase reliability headroom, reduce cold-start impact, and improve uptime consistency across all endpoints.

Short-term reliability priorities

  • Reduce perceived cold-start delays for core user flows.
  • Improve retry behavior and user-facing fallback messages.
  • Harden integrations that are more likely to fail after idle periods.
  • Keep communication clear when outages are known.

Thank you for your patience and for continuing to test, report issues, and share feedback. In the meantime, the analyzer, deck builder, and blog are all live — and the FAQ covers the most common questions.

If you want to support ongoing reliability improvements, you can buy me a beer here.

Comments

0 total

You need to sign in to comment.

Loading comments...