kubevirtci: move into it's own folder to allow OWNERS

We use [kubevirtci] for spinning up virtualized k8s clusters for
testing (locally and on CI). Therefore we source the cluster-up folder
into kubevirts' project folder root. Also we have the
cluster-up-sha.txt which is used to verify which version we use
currently, so that we can determine whether an update is necessary.

Related to how OWNERS files work, and since both of those items are
located in the root folder, we currently need a root approver for
approving pull requests that bump kubevirtci.

Thus, to be able for sig-buildsystem to own kubevirtci we move it to
it's own folder `kubevirtci`. In that folder we place an OWNERS file
with which we as sig-buildsystem claim ownership of kubevirtci.

[kubevirtci]: https://github.com/kubevirt/kubevirtci

Signed-off-by: Daniel Hiller <[email protected]>

Makefile

@@ -163,7 +163,7 @@ cluster-up:
 	./hack/cluster-up.sh
 
 cluster-down:
-	./cluster-up/down.sh
+	./kubevirtci/cluster-up/down.sh
 
 cluster-build:
 	./hack/cluster-build.sh
@@ -231,7 +231,7 @@ gofumpt:
 
 update-generated-api-testdata:
 	./hack/update-generated-api-testdata.sh
-    
+
 .PHONY: \
 	build-verify \
 	conformance \

automation/perfscale-test.sh

@@ -19,7 +19,7 @@
 
 set -ex
 
-kubectl() { KUBEVIRTCI_VERBOSE=false cluster-up/kubectl.sh "$@"; }
+kubectl() { KUBEVIRTCI_VERBOSE=false kubevirtci/cluster-up/kubectl.sh "$@"; }
 
 _prometheus_port_forward_pid=""
 trap "clean_up" EXIT SIGINT SIGTERM SIGQUIT

automation/repeated_test.sh

@@ -127,7 +127,7 @@ if (( $# > 0 )); then
 else
     # We only want to use stable providers for flake testing, thus we fetch the k8s version file from kubevirtci.
     # we stop at the first provider that is stable (aka doesn't have an rc or beta or alpha version)
-    for k8s_provider in $(cd cluster-up/cluster && ls -rd k8s-[0-9]\.[0-9][0-9]); do
+    for k8s_provider in $(cd kubevirtci/cluster-up/cluster && ls -rd k8s-[0-9]\.[0-9][0-9]); do
         # shellcheck disable=SC2154
         k8s_provider_version=$(curl --fail "https://raw.githubusercontent.com/kubevirt/kubevirtci/${kubevirtci_git_hash}/cluster-provision/k8s/${k8s_provider#"k8s-"}/version")
         if [[ "${k8s_provider_version}" =~ -(rc|alpha|beta) ]]; then
@@ -143,7 +143,7 @@ else
     fi
 fi
 echo "Test lane: ${TEST_LANE}"
-[ -d "cluster-up/cluster/${TEST_LANE}" ] || ( echo "provider ${TEST_LANE} does not exist!"; exit 1 )
+[ -d "kubevirtci/cluster-up/cluster/${TEST_LANE}" ] || ( echo "provider ${TEST_LANE} does not exist!"; exit 1 )
 
 if [[ -z ${TARGET_COMMIT-} ]]; then
     # if there's no commit provided default to the latest merge commit

automation/test.sh

@@ -108,7 +108,7 @@ if [[ $KUBEVIRT_NUM_NODES = "1" && $KUBEVIRT_INFRA_REPLICAS = "1" ]]; then
   export KUBEVIRT_DEPLOY_NFS_CSI=true
 fi
 
-if [ ! -d "cluster-up/cluster/$KUBEVIRT_PROVIDER" ]; then
+if [ ! -d "kubevirtci/cluster-up/cluster/$KUBEVIRT_PROVIDER" ]; then
   echo "The cluster provider $KUBEVIRT_PROVIDER does not exist"
   exit 1
 fi
@@ -214,8 +214,8 @@ elif [[ $TARGET =~ windows.* ]]; then
   safe_download "$WINDOWS_LOCK_PATH" "$win_image_url" "$win_image" || exit 1
 fi
 
-kubectl() { KUBEVIRTCI_VERBOSE=false cluster-up/kubectl.sh "$@"; }
-cli() { cluster-up/cli.sh "$@"; }
+kubectl() { KUBEVIRTCI_VERBOSE=false kubevirtci/cluster-up/kubectl.sh "$@"; }
+cli() { kubevirtci/cluster-up/cli.sh "$@"; }
 
 determine_cri_bin() {
     if [ "${KUBEVIRTCI_RUNTIME}" = "podman" ]; then

docs/README.md

@@ -43,14 +43,14 @@ the VMI's pod and receives lifecycle commands from virt-handler.
 
 ### Scripts
 
- * `cluster-up/kubectl.sh`: This is a wrapper around Kubernetes' kubectl command so
+ * `kubevirtci/cluster-up/kubectl.sh`: This is a wrapper around Kubernetes' kubectl command so
    that it can be run directly from this checkout without logging into a node.
- * `cluster-up/virtctl.sh` is a wrapper around `virtctl`. `virtctl` brings all
+ * `kubevirtci/cluster-up/virtctl.sh` is a wrapper around `virtctl`. `virtctl` brings all
    virtual machine specific commands with it. It is supplement to `kubectl`.
-   e.g. `cluster-up/virtctl.sh console testvm`.
- * `cluster-up/cli.sh` helps you create ephemeral kubernetes and openshift
+   e.g. `kubevirtci/cluster-up/virtctl.sh console testvm`.
+ * `kubevirtci/cluster-up/cli.sh` helps you create ephemeral kubernetes and openshift
    clusters for testing. This is helpful when direct management or access to
-   cluster nodes is necessary. e.g. `cluster-up/cli.sh ssh node01`.
+   cluster nodes is necessary. e.g. `kubevirtci/cluster-up/cli.sh ssh node01`.
 
 ### Makefile Commands
 

docs/debugging.md

@@ -1,18 +1,18 @@
 # Debugging
 
 ```bash
-cluster-up/kubectl.sh version
+kubevirtci/cluster-up/kubectl.sh version
 ```
 
 will try to connect to the apiserver.
 
 ## Retrieving Logs
 
 To investigate the logs of a pod, you can view the logs via
-`cluster-up/kubectl.sh logs`. To view the logs of `virt-api`, type
+`kubevirtci/cluster-up/kubectl.sh logs`. To view the logs of `virt-api`, type
 
 ```bash
-cluster-up/kubectl.sh logs virt-api -f
+kubevirtci/cluster-up/kubectl.sh logs virt-api -f
 ```
 
 Sometimes a container in a pod is crashlooping because of an application error
@@ -22,7 +22,7 @@ attempt, the `--previous` flag can be used. To view the logs of the container
 `virt-api` in the pod `virt-api` from the previous run, type
 
 ```bash
-cluster-up/kubectl.sh logs virt-api -f -c virt-api -p
+kubevirtci/cluster-up/kubectl.sh logs virt-api -f -c virt-api -p
 ```
 
 Note that you always have to select a container inside a pod for fetching old
@@ -33,7 +33,7 @@ logs with the `--previous` flag.
 Both, Kubernetes and KubeVirt are creating events, which can be viewed via
 
 ```bash
-cluster-up/kubectl.sh get events --all-namespaces --watch
+kubevirtci/cluster-up/kubectl.sh get events --all-namespaces --watch
 ```
 
 This way it is pretty easy to detect if a Pod or a VMI got started.
@@ -45,7 +45,7 @@ to see what is going wrong. In this case the kubectl `exec` command can be
 used. To enter `virt-api` with an interactive shell, type
 
 ```bash
-cluster-up/kubectl.sh exec virt-api -c virt-api -i -t -- sh
+kubevirtci/cluster-up/kubectl.sh exec virt-api -c virt-api -i -t -- sh
 ```
 
 ## Kubelet Logs

docs/env-providers.md

@@ -8,7 +8,7 @@ All following providers allow a common workflow:
  * `make cluster-deploy` to (re)deploy the code (no provider support needed)
  * `make cluster-sync` to build and (re)deploy the code
  * `make functests` to run the functional tests against KubeVirt
- * `cluster-up/kubectl.sh` to talk to the k8s installation
+ * `kubevirtci/cluster-up/kubectl.sh` to talk to the k8s installation
 
 It is recommended to export the `KUBEVIRT_PROVIDER` variable as part of your `.bashrc`
 file.

docs/getting-started.md

@@ -155,7 +155,7 @@ If not, try to adjust them to be the same.
 See [issue 2667](https://github.com/kubevirt/kubevirt/issues/2667)
 for more detailed info.
 ```
-# ./cluster-up/kubectl.sh get pods --all-namespaces
+# ./kubevirtci/cluster-up/kubectl.sh get pods --all-namespaces
 NAMESPACE     NAME                                      READY   STATUS             RESTARTS   AGE
 cdi           cdi-operator-5db567b486-grtk9             0/1     ImagePullBackOff   0          42m
 
@@ -200,14 +200,14 @@ Based on the used cluster, node names might be different.
 You can get the names from following command:
 
 ```bash
-# cluster-up/kubectl.sh get nodes
+# kubevirtci/cluster-up/kubectl.sh get nodes
 NAME     STATUS   ROLES                   AGE   VERSION
 node01   Ready    control-plane,worker    13s   v1.18.3
 ```
 
 Then you can execute the following command to access the node:
 ```
-# ./cluster-up/ssh.sh node01
+# ./kubevirtci/cluster-up/ssh.sh node01
 [vagrant@node01 ~]$
 ```
 
@@ -281,20 +281,20 @@ Finally start a VMI called `vmi-ephemeral`:
     # This can be done from your GIT repo, no need to log into a VMI
 
     # Create a VMI
-    ./cluster-up/kubectl.sh create -f examples/vmi-ephemeral.yaml
+    ./kubevirtci/cluster-up/kubectl.sh create -f examples/vmi-ephemeral.yaml
 
     # Sure? Let's list all created VMIs
-    ./cluster-up/kubectl.sh get vmis
+    ./kubevirtci/cluster-up/kubectl.sh get vmis
 
     # Enough, let's get rid of it
-    ./cluster-up/kubectl.sh delete -f examples/vmi-ephemeral.yaml
+    ./kubevirtci/cluster-up/kubectl.sh delete -f examples/vmi-ephemeral.yaml
 
 
     # You can actually use kubelet.sh to introspect the cluster in general
-    ./cluster-up/kubectl.sh get pods
+    ./kubevirtci/cluster-up/kubectl.sh get pods
 
     # To check the running kubevirt services you need to introspect the `kubevirt` namespace:
-    ./cluster-up/kubectl.sh -n kubevirt get pods
+    ./kubevirtci/cluster-up/kubectl.sh -n kubevirt get pods
 ```
 
 This will start a VMI on control-plane or one of the running nodes with a macvtap and a
@@ -303,18 +303,18 @@ tap networking device attached.
 #### Example
 
 ```bash
-$ ./cluster-up/kubectl.sh create -f examples/vmi-ephemeral.yaml
+$ ./kubevirtci/cluster-up/kubectl.sh create -f examples/vmi-ephemeral.yaml
 vm "vmi-ephemeral" created
 
-$ ./cluster-up/kubectl.sh get pods
+$ ./kubevirtci/cluster-up/kubectl.sh get pods
 NAME                              READY     STATUS    RESTARTS   AGE
 virt-launcher-vmi-ephemeral9q7es  1/1       Running   0          10s
 
-$ ./cluster-up/kubectl.sh get vmis
+$ ./kubevirtci/cluster-up/kubectl.sh get vmis
 NAME            AGE   PHASE     IP              NODENAME
 vmi-ephemeral   11s   Running   10.244.140.77   node02
 
-$ ./cluster-up/kubectl.sh get vmis -o json
+$ ./kubevirtci/cluster-up/kubectl.sh get vmis -o json
 {
     "kind": "List",
     "apiVersion": "v1",
@@ -352,12 +352,12 @@ to add virt-viewer installation folder to their `PATH`.
 Then, after you made sure that the VMI `vmi-ephemeral` is running, type:
 
 ```
-cluster-up/virtctl.sh vnc vmi-ephemeral
+kubevirtci/cluster-up/virtctl.sh vnc vmi-ephemeral
 ```
 
 This will start a remote session with `remote-viewer`.
 
-`cluster-up/virtctl.sh` is a wrapper around `virtctl`. `virtctl` brings all
+`kubevirtci/cluster-up/virtctl.sh` is a wrapper around `virtctl`. `virtctl` brings all
 virtual machine specific commands with it and is a supplement to `kubectl`.
 
 **Note:** If accessing your cluster through ssh, be sure to forward your X11 session in order to launch `virtctl vnc`.

docs/network/sriov.md

@@ -144,7 +144,7 @@ deploy from KubeVirt sources though.
 
 In the following example, we configure the cluster using `local` provider which
 is part of kubevirt/kubevirt repo. Please consult
-[documentation](https://github.com/kubevirt/kubevirt/blob/main/cluster-up/cluster/local/README.md)
+[documentation](https://github.com/kubevirt/kubevirt/blob/main/kubevirtci/cluster-up/cluster/local/README.md)
 for general information on setting up a host using the `local` provider.
 
 The `local` provider does not install default CNI plugins like `loopback`. So
@@ -216,8 +216,8 @@ $ go get -u -d github.com/intel/multus-cni
 $ cd $GOPATH/src/github.com/intel/multus-cni/
 $ mkdir -p /etc/cni/net.d
 $ cp images/70-multus.conf /etc/cni/net.d/
-$ ./cluster-up/kubectl.sh create -f $GOPATH/src/github.com/intel/multus-cni/images/multus-daemonset.yml
-$ ./cluster-up/kubectl.sh create -f $GOPATH/src/github.com/intel/multus-cni/images/flannel-daemonset.yml
+$ ./kubevirtci/cluster-up/kubectl.sh create -f $GOPATH/src/github.com/intel/multus-cni/images/multus-daemonset.yml
+$ ./kubevirtci/cluster-up/kubectl.sh create -f $GOPATH/src/github.com/intel/multus-cni/images/flannel-daemonset.yml
 ```
 
 Now, deploy SR-IOV device plugin. Adjust config.json file for your particular
@@ -239,20 +239,20 @@ $ cat <<EOF > /etc/pcidp/config.json
     ]
 }
 EOF
-$ ./cluster-up/kubectl.sh create -f $GOPATH/src/github.com/intel/sriov-network-device-plugin/images/sriovdp-daemonset.yaml
+$ ./kubevirtci/cluster-up/kubectl.sh create -f $GOPATH/src/github.com/intel/sriov-network-device-plugin/images/sriovdp-daemonset.yaml
 ```
 
 Deploy SR-IOV CNI plugin.
 
 ```
 $ go get -u -d github.com/intel/sriov-cni/
-$ ./cluster-up/kubectl.sh create -f $GOPATH/src/github.com/intel/sriov-cni/images/sriov-cni-daemonset.yaml
+$ ./kubevirtci/cluster-up/kubectl.sh create -f $GOPATH/src/github.com/intel/sriov-cni/images/sriov-cni-daemonset.yaml
 ```
 
 Finally, create a new SR-IOV network CRD that will use SR-IOV device plugin to allocate devices.
 
 ```
-./cluster-up/kubectl.sh create -f $GOPATH/src/github.com/intel/sriov-network-device-plugin/deployments/sriov-crd.yaml
+./kubevirtci/cluster-up/kubectl.sh create -f $GOPATH/src/github.com/intel/sriov-network-device-plugin/deployments/sriov-crd.yaml
 ```
 
 Just make sure that the network spec refers to the right resource name for

docs/software-emulation.md

@@ -29,7 +29,7 @@ Enabling software emulation is a cluster-wide setting, and is activated by
 editing the `KubeVirt` CR as follows:
 
 ```bash
-cluster-up/kubectl.sh --namespace kubevirt edit kubevirt kubevirt
+kubevirtci/cluster-up/kubectl.sh --namespace kubevirt edit kubevirt kubevirt
 ```
 
 Add the following snippet to the spec:

hack/bazel-generate.sh

@@ -10,7 +10,7 @@ rm -f vendor/libvirt.org/go/libvirt/BUILD.bazel
 # generate BUILD files
 bazel run \
     --config=${ARCHITECTURE} \
-    //:gazelle -- -exclude vendor/google.golang.org/grpc --exclude cluster-up
+    //:gazelle -- -exclude vendor/google.golang.org/grpc --exclude kubevirtci/cluster-up
 
 # inject changes to libvirt BUILD file
 bazel run \

hack/bootstrap.sh

@@ -41,7 +41,7 @@ function kubevirt::bootstrap::regenerate() {
         rm .bazeldnf/sandbox.bazelrc -f
         # Run gazelle to ensure that nogo has all build files resolved and that we can bootstrap the env.
         # This is necessary since some steps remove the vendor build files and nogo would be broken then.
-        KUBEVIRT_BOOTSTRAPPING=true bazel run --config=${ARCHITECTURE} //:gazelle -- -exclude vendor/google.golang.org/grpc --exclude cluster-up
+        KUBEVIRT_BOOTSTRAPPING=true bazel run --config=${ARCHITECTURE} //:gazelle -- -exclude vendor/google.golang.org/grpc --exclude kubevirtci/cluster-up
         KUBEVIRT_BOOTSTRAPPING=true bazel run --config ${HOST_ARCHITECTURE} //rpm:sandbox_${1}
         bazel clean
 

hack/build-verify.sh

@@ -5,17 +5,17 @@ set -e
 function report_dirty_build() {
     set +e
     echo "Build is not clean:"
-    cluster-up/virtctl.sh version
+    kubevirtci/cluster-up/virtctl.sh version
     git status
     exit 1
 }
 
 # Check that "clean" is reported at least once
-if [ -z "$(cluster-up/virtctl.sh version | grep clean)" ]; then
+if [ -z "$(kubevirtci/cluster-up/virtctl.sh version | grep clean)" ]; then
     report_dirty_build
 fi
 
 # Check that "dirty" is never reported
-if [ -n "$(cluster-up/virtctl.sh version | grep dirty)" ]; then
+if [ -n "$(kubevirtci/cluster-up/virtctl.sh version | grep dirty)" ]; then
     report_dirty_build
 fi

hack/ci/entrypoint.sh

@@ -105,7 +105,7 @@ function run_tests() {
     OC_PATH=$(get_path_or_empty_string_for_cmd oc)
     KUBECTL_PATH=$(get_path_or_empty_string_for_cmd kubectl)
 
-    tests.test -v=5 -kubeconfig=${KUBECONFIG} -container-tag=${DOCKER_TAG} -container-tag-alt= -container-prefix=${DOCKER_PREFIX} -image-prefix-alt=-kv -oc-path=${OC_PATH} -kubectl-path=${KUBECTL_PATH} -gocli-path=$(pwd)/cluster-up/cli.sh -test.timeout 420m -ginkgo.noColor -ginkgo.succinct -ginkgo.slow-spec-threshold=60s ${KUBEVIRT_TESTS_FOCUS} -junit-output=${ARTIFACT_DIR}/junit.functest.xml -installed-namespace=kubevirt -previous-release-tag= -previous-release-registry=quay.io/kubevirt -deploy-testing-infra=false
+    tests.test -v=5 -kubeconfig=${KUBECONFIG} -container-tag=${DOCKER_TAG} -container-tag-alt= -container-prefix=${DOCKER_PREFIX} -image-prefix-alt=-kv -oc-path=${OC_PATH} -kubectl-path=${KUBECTL_PATH} -gocli-path=$(pwd)/kubevirtci/cluster-up/cli.sh -test.timeout 420m -ginkgo.noColor -ginkgo.succinct -ginkgo.slow-spec-threshold=60s ${KUBEVIRT_TESTS_FOCUS} -junit-output=${ARTIFACT_DIR}/junit.functest.xml -installed-namespace=kubevirt -previous-release-tag= -previous-release-registry=quay.io/kubevirt -deploy-testing-infra=false
 }
 
 export PATH="$BIN_DIR:$PATH"

hack/cluster-build.sh

@@ -25,7 +25,7 @@ DOCKER_TAG=${DOCKER_TAG:-devel}
 DOCKER_TAG_ALT=${DOCKER_TAG_ALT:-devel_alt}
 
 source hack/common.sh
-source cluster-up/cluster/$KUBEVIRT_PROVIDER/provider.sh
+source kubevirtci/cluster-up/cluster/$KUBEVIRT_PROVIDER/provider.sh
 source hack/config.sh
 
 echo "Building ..."

hack/cluster-clean.sh

@@ -22,7 +22,7 @@ set -ex
 DOCKER_TAG=${DOCKER_TAG:-devel}
 
 source hack/common.sh
-source cluster-up/cluster/$KUBEVIRT_PROVIDER/provider.sh
+source kubevirtci/cluster-up/cluster/$KUBEVIRT_PROVIDER/provider.sh
 source hack/config.sh
 
 function patch_remove_finalizers() {

hack/cluster-deploy.sh

@@ -26,7 +26,7 @@ CDI_DV_GC=${CDI_DV_GC:--1}
 
 source hack/common.sh
 # shellcheck disable=SC1090
-source cluster-up/cluster/$KUBEVIRT_PROVIDER/provider.sh
+source kubevirtci/cluster-up/cluster/$KUBEVIRT_PROVIDER/provider.sh
 source hack/config.sh
 
 function dump_kubevirt() {
@@ -51,7 +51,7 @@ function _ensure_cdi_deployment() {
     _kubectl patch cdi ${cdi_namespace} --type merge -p '{"spec": {"config": {"insecureRegistries": [ "registry:5000", "fakeregistry:5000" ]}}}'
 
     # Configure uploadproxy override for virtctl imageupload
-    host_port=$(${KUBEVIRT_PATH}cluster-up/cli.sh ports uploadproxy | xargs)
+    host_port=$(${KUBEVIRT_PATH}kubevirtci/cluster-up/cli.sh ports uploadproxy | xargs)
     override="https://127.0.0.1:$host_port"
     _kubectl patch cdi ${cdi_namespace} --type merge -p '{"spec": {"config": {"uploadProxyURLOverride": "'"$override"'"}}}'
 

hack/cluster-patch.sh

@@ -1,7 +1,7 @@
 #!/bin/bash -xe
 
 source hack/common.sh
-source cluster-up/cluster/$KUBEVIRT_PROVIDER/provider.sh
+source kubevirtci/cluster-up/cluster/$KUBEVIRT_PROVIDER/provider.sh
 source hack/config.sh
 
 function wait_for_rollout() {

hack/cluster-sync.sh

@@ -18,6 +18,9 @@
 #
 
 set -e
+set -x
+
+source hack/config-kubevirtci.sh
 
 TEMP_FILE=$(mktemp -p /tmp -t kubevirt.deploy.XXXX)