remove minikube (#3220)

Signed-off-by: Per Goncalves da Silva <[email protected]>
Co-authored-by: Per Goncalves da Silva <[email protected]>

.gitignore

@@ -449,7 +449,6 @@ __pycache__/
 /build
 /apiserver.local.config
 e2e.namespace
-minikube.kubeconfig
 apiserver.crt
 apiserver.key
 

DEVELOPMENT.md

@@ -15,10 +15,8 @@
 
 | Requirement | install docs         |
 |-------------|----------------------|
-| Minikube    | [Minikube docs]      |
 | Kind        | [Kind docs]          |
 
-[Minikube docs]: https://minikube.sigs.k8s.io/docs/start
 [Kind docs]: https://kind.sigs.k8s.io/docs/user/quick-start
 
 ### Usage
@@ -40,7 +38,7 @@ To run the e2e tests locally:
 $ make e2e-local
 ```
 
-**NOTE:** Command `make e2e-local` supports Minikube and Kind environments. If you want to run the e2e tests on Minikube, you need to make sure Minikube is deployed in the local environment. If you want to run the e2e tests on Kind, you need to make sure Kind is deployed in the local environment and switch the kubeconfig to an existing Kind cluster.
+**NOTE:** If you want to run the e2e tests, you need to make sure Kind is deployed in the local environment and switch the kubeconfig to an existing Kind cluster.
 
 To run a specific e2e test locally:
 

doc/install/install.md

@@ -16,19 +16,6 @@ kubectl create -f deploy/upstream/quickstart/olm.yaml
 
 Check out the latest [releases on github](https://github.com/operator-framework/operator-lifecycle-manager/releases) for release-specific install instructions.
 
-## Run locally with minikube
-
-This command starts minikube, builds the OLM containers locally with the minikube-provided docker, and uses the local configuration in [local-values.yaml](local-values.yaml) to build localized deployment resources for OLM.
-
-```bash
-# To install and run locally
-$ make run-local
-```
-
-You can verify that the OLM components have been successfully deployed by running `kubectl -n olm get deployments`
-
-**NOTE** It is recommended for development purposes and will use the source locally
-
 ## OpenShift
 
 **IMPORTANT:** OLM is installed by default in OpenShift 4.0 and above.

scripts/build_bare.sh

@@ -5,12 +5,5 @@
 
 set -e
 
-if [ -z "$NO_MINIKUBE" ]; then
-  ps x | grep -q [m]inikube || minikube start --kubernetes-version="v1.11.0" --extra-config=apiserver.v=4 || { echo 'Cannot start minikube.'; exit 1; }
-  eval $(minikube docker-env) || { echo 'Cannot switch to minikube docker'; exit 1; }
-  kubectl config use-context minikube
-  umask 0077 && kubectl config view --minify --flatten --context=minikube > minikube.kubeconfig
-fi
-
 kubectl delete crds --all
 kubectl create namespace $(cat $(pwd)/e2e.namespace)

scripts/build_local.sh

@@ -6,13 +6,7 @@
 set -e
 set -o xtrace
 
-[ -x "$(command -v kind)" ] && [[ "$(kubectl config current-context)" =~ ^kind-? ]] && KIND=1 NO_MINIKUBE=1
-
-if [ -z "$NO_MINIKUBE" ]; then
-  pgrep -f "[m]inikube" >/dev/null || minikube start "${MINIKUBE_ARGS}" --extra-config=apiserver.v=4 || { echo 'Cannot start minikube.'; exit 1; }
-  eval "$(minikube docker-env)" || { echo 'Cannot switch to minikube docker'; exit 1; }
-  kubectl config use-context minikube
-fi
+[ -x "$(command -v kind)" ] && [[ "$(kubectl config current-context)" =~ ^kind-? ]] && KIND=1
 
 docker build -f e2e.Dockerfile -t quay.io/operator-framework/olm:local -t quay.io/operator-framework/olm-e2e:local ./bin
 docker build -f test/e2e/hang.Dockerfile -t hang:10 ./bin

scripts/run_e2e_bare.sh

@@ -14,4 +14,4 @@ fi
 
 echo "${test_flags}"
 go test -c -tags=bare -mod=vendor -v -o e2e-bare github.com/operator-framework/operator-lifecycle-manager/test/e2e
-./e2e-bare  -test.v -test.timeout 20m ${test_flags} -kubeconfig=${KUBECONFIG:-minikube.kubeconfig} -namespace=$(cat e2e.namespace) -olmNamespace=operator-lifecycle-manager -dummyImage=hang:10
+./e2e-bare  -test.v -test.timeout 20m ${test_flags} -kubeconfig=${KUBECONFIG:} -namespace=$(cat e2e.namespace) -olmNamespace=operator-lifecycle-manager -dummyImage=hang:10

test/e2e/README.md

@@ -4,7 +4,6 @@ This runs a series of tests against the Kubernetes API to verify that OLM is fun
 
 ## Requirements
 
-* Minikube > 0.25.0
 * Helm > 2.7.0
 
 ## How to use
@@ -55,13 +54,3 @@ make file and use `-dryRun` with `-focus` and see if the regex would trigger you
 ## Build infrastructure
 
 Note that the make file target `e2e-local` is executed by the github workflow `.github/workflows/e2e-tests.yml` and uses two parallel `go test` processes.
-
-## Running on minikube
-
-The e2e suite is also runnable on a minikube cluster. First spin up the minikube cluster manually with the desired provisioner, 
-then run `make run-local` to deploy OLM onto the cluster. Tests can be run by invoking ginkgo and passing the required command line
-arguments to the test suite. For example to run a specific test:
-
-```bash
-GO111MODULE=on go run github.com/onsi/ginkgo/v2/ginkgo -focus "static provider" -v --progress ./test/e2e -- -namespace=operators -olmNamespace=olm -dummyImage=bitnami/nginx:latest
-```