Can you confirm that your PORT variable is set to the same port your app is listening on?

sure here we go

to be honest. I never had this set. and it always worked without this env var. but this morning out of the blue the deployment didnt work anymore.

then I ran a diagnosis and it told me to set the PORT env. did it. but still am not able to successfully run the health check

Hi bro,

Setting "PORT" manually may not be enough by itself.

For a Railway healthcheck to pass, your app must actually listen on the same port and bind to "0.0.0.0", not "localhost".

Please check these things:

• Your app should listen on Railway’s "$PORT".

• Your server should bind to:

0.0.0.0

not:

localhost or 127.0.0.1

• Make sure the healthcheck path actually returns "200".

For example, if your healthcheck path is "/health", then this endpoint must respond successfully:

/health

• If you manually set "PORT", make sure the port exposed in Railway Public Networking matches the same port your app is using.

For example, if you set:

PORT=8080

then your app should listen on "8080", and Railway Public Networking should also point to port "8080".

If those ports do not match, the service can look deployed but the healthcheck or public URL may still fail.

If it worked before and suddenly failed today, it may also be worth checking Railway status or trying another region temporarily. But first, confirm the app is listening on the correct host/port and that the healthcheck endpoint returns "200".

here my current Dockerfile:

ENV PORT=8080

COPY --from=installer /app .

EXPOSE 8080

CMD ["node", "--import", "./apps/scraper/dist/instrument.js", "apps/scraper/dist/app.js", "--mode", "daemon"]

and here my application:

  app.listen(port, "::", () => {

    logger.info("Server listening", { port });

    logger.info("Health check ready", {

      url: `http://localhost:${port}/health`,

    });

    logger.info("Bull Board UI ready", {

      url: `http://localhost:${port}/admin/queues`,

    });

  });

I do not have public routing enabled on this application/service.

weird this happened out of the blue. I disabled the health check and everything seems to work properly. here the logs:

{

  "message": "Server listening",

  "severity": "info",

  "attributes": {

    "label": "scraper",

    "level": "info",

    "metadata": {

      "port": 8080

    }

  },

  "timestamp": "2026-05-25T14:21:45.456790796Z"

}

its not solved. I just disabled the health check for a brief moment to see if the application even runs

ENV PORT=8080

COPY --from=installer /app .

EXPOSE 8080

CMD ["node", "--import", "./apps/scraper/dist/instrument.js", "apps/scraper/dist/app.js", "--mode", "daemon"]

Bro, the likely issue is this line:

app.listen(port, "::", () => {

You are binding the server to IPv6 only. For Railway healthchecks, I would suggest binding to "0.0.0.0" instead:

app.listen(port, "0.0.0.0", () => {

  logger.info("Server listening", { port });

});

Also make sure "port" is coming from "process.env.PORT", with "8080" only as a fallback:

const port = process.env.PORT || 8080;

Since you do not have public routing enabled, the public networking port is not the main issue here. The important part is that the Railway healthcheck can reach the app inside the container.

So u should try:

1. Change "::" to "0.0.0.0"
2. Keep "PORT=8080"
3. Make sure "/health" returns HTTP "200"
4. Redeploy

Your Dockerfile part looks fine:

ENV PORT=8080

EXPOSE 8080

The main thing you should change first is the app bind host.