run-k8s-integration-local.sh

#!/bin/bash

set -o nounset
set -o errexit

readonly PKGDIR=${GOPATH}/src/sigs.k8s.io/gcp-compute-persistent-disk-csi-driver
readonly gke_cluster_version=${GKE_CLUSTER_VERSION:-latest}
readonly kube_version=${KUBE_VERSION:-master}
readonly test_version=${TEST_VERSION:-master}

source "${PKGDIR}/deploy/common.sh"

make -C "${PKGDIR}" test-k8s-integration

# This version of the command creates a GKE cluster. It also downloads and builds a k8s release
# so that it can run the test specified

# ${PKGDIR}/bin/k8s-integration-test --run-in-prow=false \
# --staging-image=${GCE_PD_CSI_STAGING_IMAGE} --service-account-file=${GCE_PD_SA_DIR}/cloud-sa.json \
# --deploy-overlay-name=dev --storageclass-files=sc-standard.yaml,sc-balanced.yaml,sc-ssd.yaml \
# --test-focus="External.Storage" --gce-zone="us-central1-b" \
# --deployment-strategy=gke --gke-cluster-version=${gke_cluster_version} \
# --test-version=${test_version} --num-nodes=3

# This version of the command creates a GCE cluster. It downloads and builds two k8s releases,
# one for the cluster and one for the tests, unless the cluster and test versioning is the same.

# ${PKGDIR}/bin/k8s-integration-test --run-in-prow=false \
# --staging-image=${GCE_PD_CSI_STAGING_IMAGE} --service-account-file=${GCE_PD_SA_DIR}/cloud-sa.json \
# --deploy-overlay-name=dev --storageclass-files=sc-standard.yaml,sc-balanced.yaml,sc-ssd.yaml \
# --test-focus="External.Storage" --gce-zone="us-central1-b" \
# --deployment-strategy=gce --kube-version=${kube_version} \
# --test-version=${test_version} --num-nodes=3


# This version of the command creates a regional GKE cluster. It will test with
# the latest GKE version and the master test version

# ${PKGDIR}/bin/k8s-integration-test --run-in-prow=false \
# --staging-image=${GCE_PD_CSI_STAGING_IMAGE} --service-account-file=${GCE_PD_SA_DIR}/cloud-sa.json \
# --deploy-overlay-name=dev --bringup-cluster=false --teardown-cluster=false \
# --storageclass-files=sc-standard.yaml,sc-balanced.yaml,sc-ssd.yaml --do-driver-build=false  --test-focus="schedule.a.pod.with.AllowedTopologies" \
# --gce-region="us-central1" --num-nodes=${NUM_NODES:-3} --gke-cluster-version="latest" --deployment-strategy="gke" \
# --test-version="master"

# This version of the command builds and deploys the GCE PD CSI driver.
# Points to a local K8s repository to get the e2e test binary, does not bring up
# or tear down the kubernetes cluster. In addition, it runs External Storage
# snapshot tests for the PD CSI driver.
#${PKGDIR}/bin/k8s-integration-test --run-in-prow=false \
#--staging-image=${GCE_PD_CSI_STAGING_IMAGE} --service-account-file=${GCE_PD_SA_DIR}/cloud-sa.json \
#--deploy-overlay-name=prow-gke-release-staging-head --bringup-cluster=false --teardown-cluster=false --test-focus="External.*Storage.*snapshot" --local-k8s-dir=$KTOP \
#--storageclass-files=sc-standard.yaml,sc-balanced.yaml,sc-ssd.yaml --snapshotclass-file=pd-volumesnapshotclass.yaml --do-driver-build=true \
#--gce-zone="us-central1-b" --num-nodes=${NUM_NODES:-3}

# This version of the command brings up (and subsequently tears down) a GKE
# cluster with managed GCE PersistentDisk CSI driver add-on enabled, and points to
# the local K8s repository to get the e2e test binary.
# ${PKGDIR}/bin/k8s-integration-test --run-in-prow=false --service-account-file=${GCE_PD_SA_DIR}/cloud-sa.json \
# --test-focus="External.Storage" --local-k8s-dir=$KTOP --storageclass-files=sc-standard.yaml,sc-balanced.yaml,sc-ssd.yaml \
# --snapshotclass-file=pd-volumesnapshotclass.yaml --do-driver-build=false --teardown-driver=false \
# --gce-zone="us-central1-c" --num-nodes=${NUM_NODES:-3} --gke-cluster-version="latest" --deployment-strategy="gke" \
# --use-gke-managed-driver=true --teardown-cluster=true

# This version of the command brings up (and subsequently tears down) a GKE
# cluster using a release channel, and with managed GCE PersistentDisk CSI driver add-on enabled, and points to
# the local K8s repository to get the e2e test binary.
# ${PKGDIR}/bin/k8s-integration-test --run-in-prow=false --service-account-file=${GCE_PD_SA_DIR}/cloud-sa.json \
# --test-focus="External.Storage" --local-k8s-dir=$KTOP --storageclass-files=sc-standard.yaml,sc-balanced.yaml,sc-ssd.yaml \
# --snapshotclass-file=pd-volumesnapshotclass.yaml --do-driver-build=false --teardown-driver=false \
# --gce-zone="us-central1-c" --num-nodes=${NUM_NODES:-3} --gke-release-channel="rapid" --deployment-strategy="gke" \
# --use-gke-managed-driver=true --teardown-cluster=true

# This version of the command does not build the driver or K8s, points to a local K8s repo to get 
# the e2e.test binary, does not bring up or down the cluster, and uses application default
# credentials instead of requiring a service account key.
# 
# Cluster nodes must have the proper GCP scopes set. This is done with kubetest by
# NODE_SCOPES=https://www.googleapis.com/auth/cloud-platform \
# KUBE_GCE_NODE_SERVICE_ACCOUNT=$SERVICE_ACCOUNT_NAME@$PROJECT.iam.gserviceaccount.com \
# kubetest --up
#
# GCE_PD_SA_DIR is not used.
#
# As with all other methods local credentials must be set by running 
#  gcloud auth application-default login
"${PKGDIR}/bin/k8s-integration-test" --run-in-prow=false \
--deploy-overlay-name=noauth --bringup-cluster=false --teardown-cluster=false --local-k8s-dir="$KTOP" \
--storageclass-files=sc-standard.yaml --do-driver-build=false --test-focus='External.Storage' \
--gce-zone="us-central1-b" --num-nodes="${NUM_NODES:-3}"


# This version of the command does not build the driver or K8s, points to a
# local K8s repo to get the e2e.test binary, and does not bring up or down the cluster
# "${PKGDIR}/bin/k8s-integration-test" --run-in-prow=false \
# --staging-image="${GCE_PD_CSI_STAGING_IMAGE}" --service-account-file="${GCE_PD_SA_DIR}/cloud-sa.json" \
# --deploy-overlay-name=dev --bringup-cluster=false --teardown-cluster=false --local-k8s-dir="$KTOP" \
# --storageclass-files=sc-standard.yaml --do-driver-build=false --test-focus='External.Storage' \
# --gce-zone="us-central1-b" --num-nodes="${NUM_NODES:-3}"