0

I have a load balancer for two VM instances in an instance group that load balances UDP ports 162 and 514. I created a health check for the instance group, but the health check is never firing based on my review of the logs.

The VM instances support HTTPS port 443 traffic and SSH port 22. I have tried both HTTPS / 443 and TCP / 443, as well as SSL and TCP / 22. The health check never fires.

I added the Google Cloud health check IP ranges to the ingress firewall rule for these VMs as well: 130.211.0.0/22 and 35.191.0.0/16. This had no effect.

What am I missing with my setup? I'll be happy to update this question with further details if it needs more information.

Dshiz
  • 3,099
  • 3
  • 26
  • 53
  • Are the VM instances that are part of the instance group listening to those ports 162 and 514? Did you also enable the Firewall logs? Also, please confirm if you are using an internal or external TCP/UDP Load Blancer and if its network based or target pool based. – James S May 12 '23 at 15:16
  • Please do a ``netstat -tulpn`` to see if the ports that are listening inside the VM, kindly update the question with the details asked – James S May 12 '23 at 15:18
  • The title of your post mentions `UDP`, but the Compute Engine services are using TCP. Edit your post to clarify but add details on how you configured the health check. – John Hanley May 15 '23 at 18:18

1 Answers1

0

Health check issues can come from a lot of different ways and here are the things that you should consider:

  • The target instance being monitored is going offline (Could be instance failure, network connectivity or software related issues)
  • Firewall or Network issues that prevents the health check from reaching the target instance (i.e. the ports to where the health checks are listening to are blocked)
  • Service Failures - if the application running within the instance is not working properly/returning error messages then it will be marked as unhealthy. Or if there are no existing service that are running within the instance to where the health checks are listening to.
  • Resource issues - If the target instances are experiencing high resource utilization, it may not be able to respond to the health check requests within the set time which leads to the health check rendering it unhealthy.
  • Load Balancer Configuration issues - Health check problems can also result from misconfigured forwarding rules or an incorrect health check setup, such as not listening to the correct backend service.
James S
  • 1,181
  • 1
  • 7