feat(test): creating local cluster for e2e

[onuryilmaz]

Creating a local cluster:

$ make e2e-local-cluster-create
cd test/e2e/local-cluster && go run .
I0629 13:22:52.739416  580350 create_cluster.go:63] "configuration loaded" kindClusterName="greenhouse-e2e" dockerImageRepository="greenhouse" dockerImageTag="e2e-latest" dockerImageBuildSkip=false kubeconfigName="e2e.kubeconfig" kubeconfigInternalName="e2e.internal.kubeconfig"
I0629 13:23:08.063941  580350 create_cluster.go:80] "cluster created successfully"
I0629 13:23:08.332099  580350 create_cluster.go:108] "kubeconfig exported successfully"
I0629 13:23:09.076845  580350 create_cluster.go:118] "Docker image built successfully"
I0629 13:23:16.396176  580350 create_cluster.go:127] "Docker image loaded to the cluster successfully"
I0629 13:23:39.179051  580350 create_cluster.go:135] "Greenhouse manager is deployed successfully"

[uwe-mayer]

Nice work!

Take a look at what already exists in /pkg/test namely env.go. This is what we need to use and leverage to be able to easily switch between contexts:

• env-test
• local kind
• remote greenhouse cluster

It provides a lot of tooling that we do not need to duplicate, mainly providing a client, config and kubeconfig when testing for example

[IvoGoman]

I like how this looks now.

[onuryilmaz]

⎈ /g/s/g/c/greenhouse > docker rmi greenhouse:e2e-latest
Untagged: greenhouse:e2e-latest
Deleted: sha256:0d48f04c01d1d4fec4e2c16dada00c968101b4ac5d22a8bce19e7d62409f4c71

⎈ /g/s/g/c/greenhouse > kind delete clusters greenhouse-e2e
Deleted nodes: ["greenhouse-e2e-control-plane"]
Deleted clusters: ["greenhouse-e2e"]

⎈ /g/s/g/c/greenhouse > make e2e-local-cluster-create
cd test/e2e/local-cluster && go run .
I0711 11:45:22.514215 2340085 create_cluster.go:82] "configuration loaded" kindClusterName="greenhouse-e2e" dockerImageRepository="greenhouse" dockerImageTag="e2e-latest" dockerImageBuildSkip=false kubeconfigName="e2e.kubeconfig" kubeconfigInternalName="e2e.internal.kubeconfig"
I0711 11:45:39.319226 2340085 create_cluster.go:99] "cluster created successfully"
I0711 11:45:39.555364 2340085 create_cluster.go:127] "kubeconfig exported successfully"
I0711 11:46:23.526305 2340085 create_cluster.go:137] "Docker image built successfully"
I0711 11:46:31.278263 2340085 create_cluster.go:146] "Docker image loaded to the cluster successfully"
I0711 11:46:33.850611 2340085 create_cluster.go:154] "idproxy is deployed successfully"
I0711 11:46:46.652780 2340085 create_cluster.go:162] "Greenhouse manager is deployed successfully"

⎈ /g/s/g/c/greenhouse > export KUBECONFIG=./test/e2e/local-cluster/e2e.kubeconfig
⎈ /g/s/g/c/greenhouse > kubectl get pods -n greenhouse
NAME                                            READY   STATUS      RESTARTS   AGE
greenhouse-controller-manager-bb9756c78-z48lr   1/1     Running     0          23s
greenhouse-kube-webhook-certgen-qf7dq           0/2     Completed   0          23s
⎈ /g/s/g/c/greenhouse > 

@uwe-mayer @IvoGoman fyi

[uwe-mayer]

@IvoGoman @onuryilmaz Totally agree on changing the kubeconfig env var to run remote tests against.
See 4a8f594
PR is good to go from my side...