I use Railway for infrastructure for my backend, Supabase with Postgres for my database, and Vercel for my frontend.

I just made a non-functional comment change on my main branch and pushed it, causing a redeploy for my Railway.

Suddenly, all API requests from my frontend to my backend are timing out.

However, suddenly no request from our web FE to our Railway BE are being received, and therefore are timing out on the FE.

Database is healthy

On local dev set up, using the exact same env vars production, everything works perfectly as intended.

I testing a public endpoint on my backend via browser, which is on api.hud.so. Works perfectly fine.

I hit it via a curl request, works perfectly fine.

When I hit it from my Vercel-hosted frontend Next.js web app, I never see the requests propagate through to my backend server logs, then they just time out.

Finally, I reverted the commit and now the project wont deploy.

Would love any insights here.