Nextjs crashes on Heroku roughly every 12 hours without any logs or Sentry issues

[alexanderjulmer]

Summary

I am facing the weirdest of issues: I have Nextjs running on a Heroku Dyno (Vercel's just too complicated and everything else is on Heroku as well). It works mostly fine but roughly every 12 hours the server just crashes and needs to be restarted.

Although completely configured, my server doesn't log anything and hence I cannot rely on logs for debugging.

I also have Sentry configured according to their instructions (https://docs.sentry.io/platforms/javascript/guides/nextjs/manual-setup/) but same there: no issues, nothing.

Finally, Heroku's memory monitor equally shows no issues.

The Nextjs server only serves a React application, all logic is running on a Python-based backend.

I am at a complete loss what could cause this behaviour. Your help is greatly appreciated.

Additional information

const nextConfig: NextConfig = {
  trailingSlash: true,
  pageExtensions: ["js", "jsx", "md", "mdx", "ts", "tsx"],
  images: {
    unoptimized: true,
  },
  experimental: {
    authInterrupts: true,
    serverActions: { bodySizeLimit: "5mb" },
  },
  logging: {
    fetches: {
      fullUrl: true,
    },
    incomingRequests: true,
  },
  outputFileTracingIncludes: {
    // Always include /public/i18n
    "/**": ["public/i18n/**"],
  },
}

Example

No response