Problem
You may observe the following error when launching kommander migrate command on the workstation that uses the HTTP proxy:
✗ Ensuring Traefik ingress controller is migrated traefik appdeploymnet ensured Waiting for Traefik HelmRelease to become ready with timeout 5m0s ... failed to connect to LB address lb.my-cluster.my-domain: dial tcp 192.168.1.2:443: connect: connection timed out,
will retry
10.95.3.19 LB address has been found in primary LB's certificate's SAN names
Solution
This is a known issue caused by the fact that to test the connectivity with the Traefik instance, the kommander migrate command uses a TCP connection instead of an HTTPS one.
The workaround is:
1. Forward Traefik port via kubectl to localhost:
[root@workstation]# kubectl -n kommander port-forward svc/kommander-traefik 443:443
2. Temporarily map the load-balancer address (kube-api endpoint) to localhost in /etc/hosts:
[root@workstation]# cat /etc/hosts 127.0.0.1 localhost localhost.localdomain localhost4 localhost4.localdomain4 lb.my-cluster.my-domain
3. Repeat kommander migrate command.