Forum

I just purchased a Singapore server and deployed a Node.js + Playwright service (service: flow4-playwright) on it.

I have completed multiple rounds of troubleshooting, and the conclusion is clear: the issue is not with the application itself, but with the server node / ingress / public routing layer.

【What I have confirmed is working】

1. Container internal health check is normal

• curl http://127.0.0.1:8080/health returns {"ok":true,"hasAuth":true} correctly.

2. The application itself has started normally

• Runtime Logs clearly show: flow4-playwright listening on 8080.

3. Container internal screenshot endpoint is normal

• Calling the corresponding endpoint inside the container successfully generates a real PNG image.
• Verified it is not a JSON error, but actual PNG binary.

【Abnormal Phenomena】

1. Public domain access is unstable

• 404 / 502 errors have occurred.
• The same service works normally inside the container, but the external domain cannot stably retrieve the correct content.

2. Runtime Logs show node-level errors

• NodeNotReady: Node is not ready

3. Platform health check has indicated multiple system component anomalies

• nats
• cadvisor
• fluent-bit
• ingress-controller
• node-exporter

【My Assessment】 Based on the evidence above, I judge that this is not an application configuration issue, but a problem with the node or platform network layer where this newly purchased Singapore server is located. Because:

• localhost inside the container is normal
• app process is listening normally
• functions inside the container are executable
• but public routing / domain behavior is abnormal

【Requested Action】

1. Please directly assist in confirming whether this newly purchased Singapore server is an abnormal node.
2. If it is a node/ingress issue, please assist in migrating to a healthy node or replacing it with a normal server.
3. If this server is confirmed to be abnormal, please provide credit / refund / replacement.

This server was just purchased, and I hope this issue will be treated as an infrastructure anomaly with priority, rather than just a general application deployment issue.

If needed, I can provide:

• Runtime Logs screenshots
• System health anomaly screenshots
• Container internal localhost test results
• Evidence of PNG output inside the container
• Evidence of external 404 / 502 errors

Since this is a newly purchased server and I have completed the troubleshooting of differences between inside and outside the container, I hope this case can be handled directly at the infrastructure level, rather than asking me to repeat application-layer troubleshooting.