Health Monitor Check - Automatically Change Port?

[awerhane]

I have an nginx application running on a Virtual Instance (on an Ubuntu instance). It is behind a Google Cloud Loadbalancer. It's been running for about a month and then unexpectedly this morning, it was taken out of service because it failed the Health Monitor Check. I checked the monitor and changed the port from 8080 to 80 and that resolved the issue.

I'm trying to diagnose what happened. Possible scenarios below?

1) The Health Monitor was ALWAYS port 8080 before, and nginx suddenly stopped handling requests for port 8080.

2) The Health Monitor Check automatically changed itself from port 80 to port 8080 which then created the service failure.

Is there a way to see a log of changes to a Health Check Monitor?

Could there have been an automatic update to nginx which would have changed which ports it responds to?

Any help is appreciated. Just want to prevent this from happening again.

[VannGuce]

Hi,

Upon checking your concern this is the closest documentation[1] I can share regarding the automatic change of port number in your health check wherein it was mentioned that if you omit the port specification it will use the default one.

You may use this guide[2] on how you can check the logs of your health check. There is a great possibility that you can see what is causing the automatic change of your port in health check.

[1]https://cloud.google.com/load-balancing/docs/health-checks#:~:text=If%20you%20omit%20the%20port%20sp...
[2]https://cloud.google.com/load-balancing/docs/health-check-logging#:~:text=In%20the%20Google%20Cloud%....