Great question from today:
I have a use case where I want to use a PROXY variable for deploying KOTS. I am using an external squid proxy for this purpose. It is configured for both HTTP & SSL sites.
When I try to deploy KOTS with:kubectl kots install $REPLICATED_APP \ --namespace $NAMESPACE \ --shared-password $ADMIN_PASSWORD \ --no-port-forward \ --http-proxy http://x.x.x.x:3128
Admin portal comes up but it fails to load the certificate with error:
{"level":"error","ts":"2022-08-24T23:48:54Z","msg":"failed to get latest license: failed to execute get request: Get \"https://replicated.app/license/app-enterprise\": dial tcp 162.159.134.41:443: i/o timeout"} {"level":"info","ts":"2022-08-24T23:48:54Z","msg":"method=POST status=500 duration=30.044787684s request=/api/v1/license"}
But if I deploy KOTS with explicit
--https-proxy
parameter:kubectl kots install $REPLICATED_APP \ --namespace $NAMESPACE \ --shared-password $ADMIN_PASSWORD \ --no-port-forward \ --http-proxy http://x.x.x.x:3128 \ --https-proxy https://x.x.x.x:3129
The kotsadm pod crashes with ERROR:
2022/08/24 23:41:42 Starting monitor loop W0824 23:41:42.177964 1 reflector.go:324] k8s.io/client-go@v1.23.1-k3s1/tools/cache/reflector.go:167: failed to list *v1.Job: Get "https://172.20.0.1:443/apis/batch/v1/namespaces/app-system/jobs?limit=500&resourceVersion=0": proxyconnect tcp: EOF E0824 23:41:42.178173 1 reflector.go:138] k8s.io/client-go@v1.23.1-k3s1/tools/cache/reflector.go:167: Failed to watch *v1.Job: failed to list *v1.Job: Get "https://172.20.0.1:443/apis/batch/v1/namespaces/app-system/jobs?limit=500&resourceVersion=0": proxyconnect tcp: EOF 2022/08/24 23:41:42 Shutting down the operator client panic: failed to get server groups: Get "https://172.20.0.1:443/api": proxyconnect tcp: EOF
Could you help me with few troubleshooting steps that I could try at my end to make it run.