Search code examples
google-cloud-platformgoogle-compute-engineload-balancinggoogle-cloud-http-load-balancer

Few VM instances in a backend service are unhealthy while configuring HTTPS load balancer on GCP

As part of load balancer creation using Unmanaged instance groups, we created 4 unmanaged instance groups each of them containing a single VM.

Out of the 4 VMs , 2 VMs are in the same zone europe-west1-c and are healthy and the other 2 VMs are in europe-west1-b and d zones and are unhealthy.

We have kept TCP and port 80 in the health check for the backend service. we are using a single backend service for all the 4 VMs.

Please suggest us how do we troubleshoot these unhealthy instances and fix this so that traffic can be spread across all 4 vm's.

Solution

  • Yes issue got resolved after recreating vm as it is not listening on port 80 earlier