name: Canary integration tests on: push: tags: - v* branches: - master - release-* pull_request: branches: - master - release-* defaults: run: # reference: https://docs.github.com/en/actions/reference/workflow-syntax-for-github-actions#using-a-specific-shell shell: bash --noprofile --norc -eo pipefail -x {0} jobs: canary: runs-on: ubuntu-18.04 if: "!contains(github.event.pull_request.labels.*.name, 'skip-ci')" steps: - name: checkout uses: actions/checkout@v2 with: fetch-depth: 0 - name: setup cluster resources uses: ./.github/workflows/setup-cluster-resources with: github-token: ${{ secrets.GITHUB_TOKEN }} - name: validate-yaml run: tests/scripts/github-action-helper.sh validate_yaml - name: use local disk and create partitions for osds run: | tests/scripts/github-action-helper.sh use_local_disk tests/scripts/github-action-helper.sh create_partitions_for_osds - name: deploy cluster run: tests/scripts/github-action-helper.sh deploy_cluster - name: wait for prepare pod run: tests/scripts/github-action-helper.sh wait_for_prepare_pod - name: wait for ceph to be ready run: tests/scripts/github-action-helper.sh wait_for_ceph_to_be_ready all 2 - name: test external script create-external-cluster-resources.py run: | toolbox=$(kubectl get pod -l app=rook-ceph-tools -n rook-ceph -o jsonpath='{.items[*].metadata.name}') timeout 15 sh -c "until kubectl -n rook-ceph exec $toolbox -- ceph mgr dump -f json|jq --raw-output .active_addr|grep -Eosq \"(25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?)\.(25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?)\.(25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?)\.(25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?)\" ; do sleep 1 && echo 'waiting for the manager IP to be available'; done" mgr_raw=$(kubectl -n rook-ceph exec $toolbox -- ceph mgr dump -f json|jq --raw-output .active_addr) timeout 60 sh -c "until kubectl -n rook-ceph exec $toolbox -- curl --silent --show-error ${mgr_raw%%:*}:9283; do echo 'waiting for mgr prometheus exporter to be ready' && sleep 1; done" kubectl -n rook-ceph exec $toolbox -- mkdir -p /etc/ceph/test-data kubectl -n rook-ceph cp tests/ceph-status-out $toolbox:/etc/ceph/test-data/ kubectl -n rook-ceph cp deploy/examples/create-external-cluster-resources.py $toolbox:/etc/ceph timeout 10 sh -c "until kubectl -n rook-ceph exec $toolbox -- python3 /etc/ceph/create-external-cluster-resources.py --rbd-data-pool-name replicapool; do echo 'waiting for script to succeed' && sleep 1; done" - name: run external script create-external-cluster-resources.py unit tests run: | kubectl -n rook-ceph exec $(kubectl get pod -l app=rook-ceph-tools -n rook-ceph -o jsonpath='{.items[0].metadata.name}') -- python3 -m unittest /etc/ceph/create-external-cluster-resources.py # write a test file # copy the test file # execute the test file - name: check-ownerreferences run: tests/scripts/github-action-helper.sh check_ownerreferences - name: test osd removal jobs run: | kubectl -n rook-ceph delete deploy/rook-ceph-operator kubectl -n rook-ceph delete deploy/rook-ceph-osd-1 --grace-period=0 --force sed -i 's//1/' deploy/examples/osd-purge.yaml # the CI must force the deletion since we use replica 1 on 2 OSDs sed -i 's/false/true/' deploy/examples/osd-purge.yaml sed -i 's|rook/ceph:master|rook/ceph:local-build|' deploy/examples/osd-purge.yaml kubectl -n rook-ceph create -f deploy/examples/osd-purge.yaml toolbox=$(kubectl get pod -l app=rook-ceph-tools -n rook-ceph -o jsonpath='{.items[*].metadata.name}') kubectl -n rook-ceph exec $toolbox -- ceph status timeout 120 sh -c "until kubectl -n rook-ceph exec $toolbox -- ceph osd tree|grep -qE 'osd.1.*.destroyed'; do echo 'waiting for ceph osd 1 to be destroyed'; sleep 1; done" kubectl -n rook-ceph exec $toolbox -- ceph status kubectl -n rook-ceph exec $toolbox -- ceph osd tree - name: collect common logs if: always() uses: ./.github/workflows/collect-logs with: name: canary - name: setup tmate session for debugging when event is PR if: failure() && github.event_name == 'pull_request' uses: mxschmitt/action-tmate@v3 timeout-minutes: 60 pvc: runs-on: ubuntu-18.04 if: "!contains(github.event.pull_request.labels.*.name, 'skip-ci')" steps: - name: checkout uses: actions/checkout@v2 with: fetch-depth: 0 - name: setup cluster resources uses: ./.github/workflows/setup-cluster-resources with: github-token: ${{ secrets.GITHUB_TOKEN }} - name: use local disk and create partitions for osds run: | tests/scripts/github-action-helper.sh use_local_disk tests/scripts/github-action-helper.sh create_partitions_for_osds - name: create cluster prerequisites run: | BLOCK=$(sudo lsblk --paths|awk '/14G/ {print $1}'| head -1) tests/scripts/localPathPV.sh "$BLOCK" tests/scripts/github-action-helper.sh create_cluster_prerequisites - name: deploy cluster run: | tests/scripts/github-action-helper.sh deploy_manifest_with_local_build deploy/examples/operator.yaml yq write -i tests/manifests/test-cluster-on-pvc-encrypted.yaml "spec.storage.storageClassDeviceSets[0].encrypted" false yq write -i tests/manifests/test-cluster-on-pvc-encrypted.yaml "spec.storage.storageClassDeviceSets[0].count" 2 yq write -i tests/manifests/test-cluster-on-pvc-encrypted.yaml "spec.storage.storageClassDeviceSets[0].volumeClaimTemplates[0].spec.resources.requests.storage" 6Gi kubectl create -f tests/manifests/test-cluster-on-pvc-encrypted.yaml tests/scripts/github-action-helper.sh deploy_manifest_with_local_build deploy/examples/toolbox.yaml - name: wait for prepare pod run: tests/scripts/github-action-helper.sh wait_for_prepare_pod - name: wait for ceph to be ready run: tests/scripts/github-action-helper.sh wait_for_ceph_to_be_ready osd 2 - name: check-ownerreferences run: tests/scripts/github-action-helper.sh check_ownerreferences - name: collect common logs if: always() uses: ./.github/workflows/collect-logs with: name: pvc - name: setup tmate session for debugging when event is PR if: failure() && github.event_name == 'pull_request' uses: mxschmitt/action-tmate@v3 timeout-minutes: 60 pvc-db: runs-on: ubuntu-18.04 if: "!contains(github.event.pull_request.labels.*.name, 'skip-ci')" steps: - name: checkout uses: actions/checkout@v2 with: fetch-depth: 0 - name: setup cluster resources uses: ./.github/workflows/setup-cluster-resources with: github-token: ${{ secrets.GITHUB_TOKEN }} - name: use local disk run: tests/scripts/github-action-helper.sh use_local_disk - name: create bluestore partitions and PVCs run: tests/scripts/github-action-helper.sh create_bluestore_partitions_and_pvcs - name: create cluster prerequisites run: tests/scripts/github-action-helper.sh create_cluster_prerequisites - name: deploy cluster run: | tests/scripts/github-action-helper.sh deploy_manifest_with_local_build deploy/examples/operator.yaml yq write -i tests/manifests/test-cluster-on-pvc-encrypted.yaml "spec.storage.storageClassDeviceSets[0].encrypted" false cat tests/manifests/test-on-pvc-db.yaml >> tests/manifests/test-cluster-on-pvc-encrypted.yaml kubectl create -f tests/manifests/test-cluster-on-pvc-encrypted.yaml tests/scripts/github-action-helper.sh deploy_manifest_with_local_build deploy/examples/toolbox.yaml - name: wait for prepare pod run: tests/scripts/github-action-helper.sh wait_for_prepare_pod - name: wait for ceph to be ready run: tests/scripts/github-action-helper.sh wait_for_ceph_to_be_ready osd 1 - name: collect common logs if: always() uses: ./.github/workflows/collect-logs with: name: pvc-db - name: setup tmate session for debugging when event is PR if: failure() && github.event_name == 'pull_request' uses: mxschmitt/action-tmate@v3 timeout-minutes: 60 pvc-db-wal: runs-on: ubuntu-18.04 if: "!contains(github.event.pull_request.labels.*.name, 'skip-ci')" steps: - name: checkout uses: actions/checkout@v2 with: fetch-depth: 0 - name: setup cluster resources uses: ./.github/workflows/setup-cluster-resources with: github-token: ${{ secrets.GITHUB_TOKEN }} - name: use local disk run: tests/scripts/github-action-helper.sh use_local_disk - name: create bluestore partitions and PVCs for wal run: tests/scripts/github-action-helper.sh create_bluestore_partitions_and_pvcs_for_wal - name: create cluster prerequisites run: tests/scripts/github-action-helper.sh create_cluster_prerequisites - name: deploy rook run: | tests/scripts/github-action-helper.sh deploy_manifest_with_local_build deploy/examples/operator.yaml yq write -i tests/manifests/test-cluster-on-pvc-encrypted.yaml "spec.storage.storageClassDeviceSets[0].encrypted" false cat tests/manifests/test-on-pvc-db.yaml >> tests/manifests/test-cluster-on-pvc-encrypted.yaml cat tests/manifests/test-on-pvc-wal.yaml >> tests/manifests/test-cluster-on-pvc-encrypted.yaml kubectl create -f tests/manifests/test-cluster-on-pvc-encrypted.yaml tests/scripts/github-action-helper.sh deploy_manifest_with_local_build deploy/examples/toolbox.yaml - name: wait for prepare pod run: tests/scripts/github-action-helper.sh wait_for_prepare_pod - name: wait for ceph to be ready run: | tests/scripts/github-action-helper.sh wait_for_ceph_to_be_ready osd 1 kubectl -n rook-ceph get pods - name: collect common logs if: always() uses: ./.github/workflows/collect-logs with: name: pvc-db-wal - name: setup tmate session for debugging when event is PR if: failure() && github.event_name == 'pull_request' uses: mxschmitt/action-tmate@v3 timeout-minutes: 60 encryption-pvc: runs-on: ubuntu-18.04 if: "!contains(github.event.pull_request.labels.*.name, 'skip-ci')" steps: - name: checkout uses: actions/checkout@v2 with: fetch-depth: 0 - name: setup cluster resources uses: ./.github/workflows/setup-cluster-resources with: github-token: ${{ secrets.GITHUB_TOKEN }} - name: use local disk and create partitions for osds run: | tests/scripts/github-action-helper.sh use_local_disk tests/scripts/github-action-helper.sh create_partitions_for_osds - name: create cluster prerequisites run: | tests/scripts/localPathPV.sh $(lsblk --paths|awk '/14G/ {print $1}'| head -1) tests/scripts/github-action-helper.sh create_cluster_prerequisites - name: deploy cluster run: | tests/scripts/github-action-helper.sh deploy_manifest_with_local_build deploy/examples/operator.yaml yq write -i tests/manifests/test-cluster-on-pvc-encrypted.yaml "spec.storage.storageClassDeviceSets[0].count" 2 yq write -i tests/manifests/test-cluster-on-pvc-encrypted.yaml "spec.storage.storageClassDeviceSets[0].volumeClaimTemplates[0].spec.resources.requests.storage" 6Gi kubectl create -f tests/manifests/test-cluster-on-pvc-encrypted.yaml tests/scripts/github-action-helper.sh deploy_manifest_with_local_build deploy/examples/toolbox.yaml - name: wait for prepare pod run: tests/scripts/github-action-helper.sh wait_for_prepare_pod - name: wait for ceph to be ready run: | tests/scripts/github-action-helper.sh wait_for_ceph_to_be_ready osd 2 kubectl -n rook-ceph get secrets sudo lsblk - name: collect common logs if: always() uses: ./.github/workflows/collect-logs with: name: encryption-pvc - name: setup tmate session for debugging when event is PR if: failure() && github.event_name == 'pull_request' uses: mxschmitt/action-tmate@v3 timeout-minutes: 60 encryption-pvc-db: runs-on: ubuntu-18.04 if: "!contains(github.event.pull_request.labels.*.name, 'skip-ci')" steps: - name: checkout uses: actions/checkout@v2 with: fetch-depth: 0 - name: setup cluster resources uses: ./.github/workflows/setup-cluster-resources with: github-token: ${{ secrets.GITHUB_TOKEN }} - name: use local disk run: tests/scripts/github-action-helper.sh use_local_disk - name: create bluestore partitions and PVCs run: tests/scripts/github-action-helper.sh create_bluestore_partitions_and_pvcs - name: create cluster prerequisites run: tests/scripts/github-action-helper.sh create_cluster_prerequisites - name: deploy cluster run: | tests/scripts/github-action-helper.sh deploy_manifest_with_local_build deploy/examples/operator.yaml cat tests/manifests/test-on-pvc-db.yaml >> tests/manifests/test-cluster-on-pvc-encrypted.yaml kubectl create -f tests/manifests/test-cluster-on-pvc-encrypted.yaml tests/scripts/github-action-helper.sh deploy_manifest_with_local_build deploy/examples/toolbox.yaml - name: wait for prepare pod run: tests/scripts/github-action-helper.sh wait_for_prepare_pod - name: wait for ceph to be ready run: | tests/scripts/github-action-helper.sh wait_for_ceph_to_be_ready osd 1 kubectl -n rook-ceph get pods kubectl -n rook-ceph get secrets - name: collect common logs if: always() uses: ./.github/workflows/collect-logs with: name: encryption-pvc-db - name: setup tmate session for debugging when event is PR if: failure() && github.event_name == 'pull_request' uses: mxschmitt/action-tmate@v3 timeout-minutes: 60 encryption-pvc-db-wal: runs-on: ubuntu-18.04 if: "!contains(github.event.pull_request.labels.*.name, 'skip-ci')" steps: - name: checkout uses: actions/checkout@v2 with: fetch-depth: 0 - name: setup cluster resources uses: ./.github/workflows/setup-cluster-resources with: github-token: ${{ secrets.GITHUB_TOKEN }} - name: use local disk run: tests/scripts/github-action-helper.sh use_local_disk - name: create bluestore partitions and PVCs for wal run: tests/scripts/github-action-helper.sh create_bluestore_partitions_and_pvcs_for_wal - name: create cluster prerequisites run: tests/scripts/github-action-helper.sh create_cluster_prerequisites - name: deploy rook run: | tests/scripts/github-action-helper.sh deploy_manifest_with_local_build deploy/examples/operator.yaml cat tests/manifests/test-on-pvc-db.yaml >> tests/manifests/test-cluster-on-pvc-encrypted.yaml cat tests/manifests/test-on-pvc-wal.yaml >> tests/manifests/test-cluster-on-pvc-encrypted.yaml kubectl create -f tests/manifests/test-cluster-on-pvc-encrypted.yaml tests/scripts/github-action-helper.sh deploy_manifest_with_local_build deploy/examples/toolbox.yaml - name: wait for prepare pod run: tests/scripts/github-action-helper.sh wait_for_prepare_pod - name: wait for ceph to be ready run: | tests/scripts/github-action-helper.sh wait_for_ceph_to_be_ready osd 1 kubectl -n rook-ceph get pods kubectl -n rook-ceph get secrets - name: collect common logs if: always() uses: ./.github/workflows/collect-logs with: name: encryption-pvc-db-wal - name: setup tmate session for debugging when event is PR if: failure() && github.event_name == 'pull_request' uses: mxschmitt/action-tmate@v3 timeout-minutes: 60 encryption-pvc-kms-vault-token-auth: runs-on: ubuntu-18.04 if: "!contains(github.event.pull_request.labels.*.name, 'skip-ci')" steps: - name: checkout uses: actions/checkout@v2 with: fetch-depth: 0 - name: setup cluster resources uses: ./.github/workflows/setup-cluster-resources with: github-token: ${{ secrets.GITHUB_TOKEN }} - name: use local disk and create partitions for osds run: | tests/scripts/github-action-helper.sh use_local_disk tests/scripts/github-action-helper.sh create_partitions_for_osds - name: create cluster prerequisites run: | tests/scripts/localPathPV.sh $(lsblk --paths|awk '/14G/ {print $1}'| head -1) tests/scripts/github-action-helper.sh create_cluster_prerequisites - name: deploy vault run: tests/scripts/deploy-validate-vault.sh deploy - name: deploy cluster run: | tests/scripts/github-action-helper.sh deploy_manifest_with_local_build deploy/examples/operator.yaml cat tests/manifests/test-kms-vault.yaml >> tests/manifests/test-cluster-on-pvc-encrypted.yaml yq merge --inplace --arrays append tests/manifests/test-cluster-on-pvc-encrypted.yaml tests/manifests/test-kms-vault-spec-token-auth.yaml yq write -i tests/manifests/test-cluster-on-pvc-encrypted.yaml "spec.storage.storageClassDeviceSets[0].count" 2 yq write -i tests/manifests/test-cluster-on-pvc-encrypted.yaml "spec.storage.storageClassDeviceSets[0].volumeClaimTemplates[0].spec.resources.requests.storage" 6Gi kubectl create -f tests/manifests/test-cluster-on-pvc-encrypted.yaml yq merge --inplace --arrays append tests/manifests/test-object.yaml tests/manifests/test-kms-vault-spec-token-auth.yaml yq write -i tests/manifests/test-object.yaml "spec.security.kms.connectionDetails.VAULT_BACKEND_PATH" rook/ver2 kubectl create -f tests/manifests/test-object.yaml tests/scripts/github-action-helper.sh deploy_manifest_with_local_build deploy/examples/toolbox.yaml - name: wait for prepare pod run: tests/scripts/github-action-helper.sh wait_for_prepare_pod - name: wait for ceph to be ready run: | tests/scripts/github-action-helper.sh wait_for_ceph_to_be_ready osd 2 tests/scripts/validate_cluster.sh rgw kubectl -n rook-ceph get pods kubectl -n rook-ceph get secrets - name: validate osd vault run: | tests/scripts/deploy-validate-vault.sh validate_osd sudo lsblk - name: validate rgw vault run: | tests/scripts/deploy-validate-vault.sh validate_rgw - name: collect common logs if: always() uses: ./.github/workflows/collect-logs with: name: encryption-pvc-kms-vault-token-auth - name: setup tmate session for debugging when event is PR if: failure() && github.event_name == 'pull_request' uses: mxschmitt/action-tmate@v3 timeout-minutes: 60 encryption-pvc-kms-vault-k8s-auth: runs-on: ubuntu-18.04 if: "!contains(github.event.pull_request.labels.*.name, 'skip-ci')" steps: - name: checkout uses: actions/checkout@v2 with: fetch-depth: 0 - name: setup cluster resources uses: ./.github/workflows/setup-cluster-resources with: github-token: ${{ secrets.GITHUB_TOKEN }} - name: use local disk and create partitions for osds run: | tests/scripts/github-action-helper.sh use_local_disk tests/scripts/github-action-helper.sh create_partitions_for_osds - name: create cluster prerequisites run: | tests/scripts/localPathPV.sh $(lsblk --paths|awk '/14G/ {print $1}'| head -1) tests/scripts/github-action-helper.sh create_cluster_prerequisites - name: deploy vault run: KUBERNETES_AUTH=true tests/scripts/deploy-validate-vault.sh deploy - name: deploy cluster run: | tests/scripts/github-action-helper.sh deploy_manifest_with_local_build deploy/examples/operator.yaml yq merge --inplace --arrays append tests/manifests/test-cluster-on-pvc-encrypted.yaml tests/manifests/test-kms-vault-spec-k8s-auth.yaml yq write -i tests/manifests/test-cluster-on-pvc-encrypted.yaml "spec.storage.storageClassDeviceSets[0].count" 2 yq write -i tests/manifests/test-cluster-on-pvc-encrypted.yaml "spec.storage.storageClassDeviceSets[0].volumeClaimTemplates[0].spec.resources.requests.storage" 6Gi kubectl create -f tests/manifests/test-cluster-on-pvc-encrypted.yaml tests/scripts/github-action-helper.sh deploy_manifest_with_local_build deploy/examples/toolbox.yaml - name: wait for prepare pod run: tests/scripts/github-action-helper.sh wait_for_prepare_pod - name: wait for ceph to be ready run: | tests/scripts/github-action-helper.sh wait_for_ceph_to_be_ready osd 2 kubectl -n rook-ceph get pods kubectl -n rook-ceph get secrets - name: validate osd vault run: | tests/scripts/deploy-validate-vault.sh validate_osd sudo lsblk - name: collect common logs if: always() uses: ./.github/workflows/collect-logs with: name: encryption-pvc-kms-vault-k8s-auth - name: setup tmate session for debugging when event is PR if: failure() && github.event_name == 'pull_request' uses: mxschmitt/action-tmate@v3 timeout-minutes: 60 lvm-pvc: runs-on: ubuntu-18.04 if: "!contains(github.event.pull_request.labels.*.name, 'skip-ci')" steps: - name: checkout uses: actions/checkout@v2 with: fetch-depth: 0 - name: setup cluster resources uses: ./.github/workflows/setup-cluster-resources with: github-token: ${{ secrets.GITHUB_TOKEN }} - name: use local disk run: tests/scripts/github-action-helper.sh use_local_disk - name: create LV on disk run: tests/scripts/github-action-helper.sh create_LV_on_disk - name: deploy cluster run: | tests/scripts/github-action-helper.sh deploy_manifest_with_local_build deploy/examples/operator.yaml yq write -i tests/manifests/test-cluster-on-pvc-encrypted.yaml "spec.storage.storageClassDeviceSets[0].encrypted" false kubectl create -f tests/manifests/test-cluster-on-pvc-encrypted.yaml tests/scripts/github-action-helper.sh deploy_manifest_with_local_build deploy/examples/toolbox.yaml - name: wait for prepare pod run: tests/scripts/github-action-helper.sh wait_for_prepare_pod - name: wait for ceph to be ready run: tests/scripts/github-action-helper.sh wait_for_ceph_to_be_ready osd 1 - name: check-ownerreferences run: tests/scripts/github-action-helper.sh check_ownerreferences - name: collect common logs if: always() uses: ./.github/workflows/collect-logs with: name: lvm-pvc - name: setup tmate session for debugging when event is PR if: failure() && github.event_name == 'pull_request' uses: mxschmitt/action-tmate@v3 timeout-minutes: 60 multi-cluster-mirroring: runs-on: ubuntu-18.04 if: "!contains(github.event.pull_request.labels.*.name, 'skip-ci')" steps: - name: checkout uses: actions/checkout@v2 with: fetch-depth: 0 - name: setup cluster resources uses: ./.github/workflows/setup-cluster-resources with: github-token: ${{ secrets.GITHUB_TOKEN }} - name: use local disk into two partitions run: | tests/scripts/github-action-helper.sh use_local_disk BLOCK=$(sudo lsblk --paths|awk '/14G/ {print $1}'| head -1) tests/scripts/create-bluestore-partitions.sh --disk "$BLOCK" --osd-count 2 sudo lsblk - name: deploy first cluster rook run: | tests/scripts/github-action-helper.sh deploy_first_rook_cluster cd deploy/examples/ kubectl create -f rbdmirror.yaml -f filesystem-mirror.yaml # cephfs-mirroring is a push operation # running bootstrap create on secondary and bootstrap import on primary. mirror daemons on primary. - name: deploy second cluster rook run: | tests/scripts/github-action-helper.sh deploy_second_rook_cluster cd deploy/examples/ sed -i 's/namespace: rook-ceph/namespace: rook-ceph-secondary/g' rbdmirror.yaml kubectl create -f rbdmirror.yaml - name: wait for ceph cluster 1 to be ready run: | mkdir test tests/scripts/validate_cluster.sh osd 1 kubectl -n rook-ceph get pods - name: create replicated mirrored pool on cluster 1 run: | cd deploy/examples/ yq w -i pool-test.yaml spec.mirroring.enabled true yq w -i pool-test.yaml spec.mirroring.mode image kubectl create -f pool-test.yaml timeout 60 sh -c 'until [ "$(kubectl -n rook-ceph get cephblockpool replicapool -o jsonpath='{.status.phase}'|grep -c "Ready")" -eq 1 ]; do echo "waiting for pool replicapool to created on cluster 1" && sleep 1; done' - name: create replicated mirrored pool 2 on cluster 1 run: | cd deploy/examples/ yq w -i pool-test.yaml metadata.name replicapool2 kubectl create -f pool-test.yaml timeout 60 sh -c 'until [ "$(kubectl -n rook-ceph get cephblockpool replicapool2 -o jsonpath='{.status.phase}'|grep -c "Ready")" -eq 1 ]; do echo "waiting for pool replicapool2 to created on cluster 2" && sleep 1; done' yq w -i pool-test.yaml metadata.name replicapool - name: create replicated mirrored pool on cluster 2 run: | cd deploy/examples/ yq w -i pool-test.yaml metadata.namespace rook-ceph-secondary kubectl create -f pool-test.yaml timeout 60 sh -c 'until [ "$(kubectl -n rook-ceph-secondary get cephblockpool replicapool -o jsonpath='{.status.phase}'|grep -c "Ready")" -eq 1 ]; do echo "waiting for pool replicapool to created on cluster 1" && sleep 1; done' - name: create replicated mirrored pool 2 on cluster 2 run: | cd deploy/examples/ yq w -i pool-test.yaml metadata.name replicapool2 kubectl create -f pool-test.yaml timeout 60 sh -c 'until [ "$(kubectl -n rook-ceph-secondary get cephblockpool replicapool -o jsonpath='{.status.phase}'|grep -c "Ready")" -eq 1 ]; do echo "waiting for pool replicapool2 to created on cluster 2" && sleep 1; done' - name: create images in the pools run: | kubectl exec -n rook-ceph deploy/rook-ceph-tools -ti -- rbd -p replicapool create test -s 1G kubectl exec -n rook-ceph deploy/rook-ceph-tools -t -- rbd mirror image enable replicapool/test snapshot kubectl exec -n rook-ceph deploy/rook-ceph-tools -t -- rbd -p replicapool info test kubectl exec -n rook-ceph deploy/rook-ceph-tools -ti -- rbd -p replicapool2 create test -s 1G kubectl exec -n rook-ceph deploy/rook-ceph-tools -t -- rbd mirror image enable replicapool2/test snapshot kubectl exec -n rook-ceph deploy/rook-ceph-tools -t -- rbd -p replicapool2 info test - name: copy block mirror peer secret into the other cluster for replicapool run: | kubectl -n rook-ceph get secret pool-peer-token-replicapool -o yaml > pool-peer-token-replicapool.yaml yq delete --inplace pool-peer-token-replicapool.yaml metadata.ownerReferences yq write --inplace pool-peer-token-replicapool.yaml metadata.namespace rook-ceph-secondary yq write --inplace pool-peer-token-replicapool.yaml metadata.name pool-peer-token-replicapool-config kubectl create --namespace=rook-ceph-secondary -f pool-peer-token-replicapool.yaml - name: copy block mirror peer secret into the other cluster for replicapool2 (using cluster global peer) run: | kubectl -n rook-ceph get secret cluster-peer-token-my-cluster -o yaml > cluster-peer-token-my-cluster.yaml yq delete --inplace cluster-peer-token-my-cluster.yaml metadata.ownerReferences yq write --inplace cluster-peer-token-my-cluster.yaml metadata.namespace rook-ceph-secondary yq write --inplace cluster-peer-token-my-cluster.yaml metadata.name cluster-peer-token-my-cluster-config kubectl create --namespace=rook-ceph-secondary -f cluster-peer-token-my-cluster.yaml - name: add block mirror peer secret to the other cluster for replicapool run: | kubectl -n rook-ceph-secondary patch cephblockpool replicapool --type merge -p '{"spec":{"mirroring":{"peers": {"secretNames": ["pool-peer-token-replicapool-config"]}}}}' - name: add block mirror peer secret to the other cluster for replicapool2 (using cluster global peer) run: | kubectl -n rook-ceph-secondary patch cephblockpool replicapool2 --type merge -p '{"spec":{"mirroring":{"peers": {"secretNames": ["cluster-peer-token-my-cluster-config"]}}}}' - name: verify image has been mirrored for replicapool run: | # let's wait a bit for the image to be present timeout 120 sh -c 'until [ "$(kubectl exec -n rook-ceph-secondary deploy/rook-ceph-tools -t -- rbd -p replicapool ls|grep -c test)" -eq 1 ]; do echo "waiting for image to be mirrored in pool replicapool" && sleep 1; done' - name: verify image has been mirrored for replicapool2 run: | # let's wait a bit for the image to be present timeout 120 sh -c 'until [ "$(kubectl exec -n rook-ceph-secondary deploy/rook-ceph-tools -t -- rbd -p replicapool2 ls|grep -c test)" -eq 1 ]; do echo "waiting for image to be mirrored in pool replicapool2" && sleep 1; done' - name: display cephblockpool and image status run: | timeout 80 sh -c 'until [ "$(kubectl -n rook-ceph-secondary get cephblockpool replicapool -o jsonpath='{.status.mirroringStatus.summary.daemon_health}'|grep -c OK)" -eq 1 ]; do echo "waiting for mirroring status to be updated in replicapool" && sleep 1; done' timeout 80 sh -c 'until [ "$(kubectl -n rook-ceph-secondary get cephblockpool replicapool2 -o jsonpath='{.status.mirroringStatus.summary.daemon_health}'|grep -c OK)" -eq 1 ]; do echo "waiting for mirroring status to be updated in replicapool2" && sleep 1; done' kubectl -n rook-ceph-secondary get cephblockpool replicapool -o yaml kubectl -n rook-ceph-secondary get cephblockpool replicapool2 -o yaml kubectl exec -n rook-ceph deploy/rook-ceph-tools -t -- rbd -p replicapool info test kubectl exec -n rook-ceph deploy/rook-ceph-tools -t -- rbd -p replicapool2 info test - name: copy block mirror peer secret into the primary cluster for replicapool run: | kubectl -n rook-ceph-secondary get secret pool-peer-token-replicapool -o yaml |\ sed 's/namespace: rook-ceph-secondary/namespace: rook-ceph/g; s/name: pool-peer-token-replicapool/name: pool-peer-token-replicapool-config/g' |\ kubectl create --namespace=rook-ceph -f - - name: add block mirror peer secret to the primary cluster for replicapool run: | kubectl -n rook-ceph patch cephblockpool replicapool --type merge -p '{"spec":{"mirroring":{"peers": {"secretNames": ["pool-peer-token-replicapool-config"]}}}}' - name: wait for rook-ceph-csi-mapping-config to be updated with cluster ID run: | timeout 60 sh -c 'until [ "$(kubectl get cm -n rook-ceph rook-ceph-csi-mapping-config -o jsonpath='{.data.csi-mapping-config-json}' | grep -c "rook-ceph-secondary")" -eq 1 ]; do echo "waiting for rook-ceph-csi-mapping-config to be created with cluster ID mappings" && sleep 1; done' - name: create replicated mirrored filesystem on cluster 1 run: | PRIMARY_YAML=deploy/examples/filesystem-test-primary.yaml cp deploy/examples/filesystem-test.yaml "$PRIMARY_YAML" yq merge --inplace --arrays append "$PRIMARY_YAML" tests/manifests/test-fs-mirror-spec.yaml kubectl create -f "$PRIMARY_YAML" timeout 120 sh -c 'until [ "$(kubectl -n rook-ceph get cephfilesystem myfs -o jsonpath='{.status.phase}'|grep -c "Ready")" -eq 1 ]; do echo "waiting for filesystem to be created" && sleep 1; done' - name: create replicated mirrored filesystem on cluster 2 run: | cd deploy/examples/ yq w -i filesystem-test.yaml metadata.namespace rook-ceph-secondary yq w -i filesystem-test.yaml spec.mirroring.enabled true kubectl create -f filesystem-test.yaml timeout 120 sh -c 'until [ "$(kubectl -n rook-ceph-secondary get cephfilesystem myfs -o jsonpath='{.status.phase}'|grep -c "Ready")" -eq 1 ]; do echo "waiting for filesystem to be created" && sleep 1; done' - name: copy filesystem mirror peer secret from the secondary cluster to the primary one run: | kubectl -n rook-ceph-secondary get secret fs-peer-token-myfs -o yaml |\ sed '/ownerReferences/,+6d' |\ sed 's/namespace: rook-ceph-secondary/namespace: rook-ceph/g; s/name: fs-peer-token-myfs/name: fs-peer-token-myfs-config/g' |\ kubectl create --namespace=rook-ceph -f - - name: add filesystem mirror peer secret to the primary cluster run: | kubectl -n rook-ceph patch cephfilesystem myfs --type merge -p '{"spec":{"mirroring":{"peers": {"secretNames": ["fs-peer-token-myfs-config"]}}}}' - name: verify fs mirroring is working run: | timeout 45 sh -c 'until [ "$(kubectl -n rook-ceph exec -t deploy/rook-ceph-fs-mirror -- ls -1 /var/run/ceph/|grep -c asok)" -lt 3 ]; do echo "waiting for connection to peer" && sleep 1; done' sockets=$(kubectl -n rook-ceph exec -t deploy/rook-ceph-fs-mirror -- ls -1 /var/run/ceph/) status=$(for socket in $sockets; do minikube kubectl -- -n rook-ceph exec -t deploy/rook-ceph-fs-mirror -- ceph --admin-daemon /var/run/ceph/$socket help|awk -F ":" '/get filesystem mirror status/ {print $1}'; done) if [ "${#status}" -lt 1 ]; then echo "peer addition failed" && exit 1; fi - name: display cephfilesystem and fs mirror daemon status run: | kubectl -n rook-ceph get cephfilesystem myfs -o yaml # the check is not super ideal since 'mirroring_failed' is only displayed when there is a failure but not when it's working... timeout 60 sh -c 'while [ "$(kubectl exec -n rook-ceph deploy/rook-ceph-tools -t -- ceph fs snapshot mirror daemon status myfs|jq -r '.[0].filesystems[0]'|grep -c "mirroring_failed")" -eq 1 ]; do echo "waiting for filesystem to be mirrored" && sleep 1; done' - name: collect common logs if: always() uses: ./.github/workflows/collect-logs with: name: multi-cluster-mirroring - name: setup tmate session for debugging when event is PR if: failure() && github.event_name == 'pull_request' uses: mxschmitt/action-tmate@v3 timeout-minutes: 60 rgw-multisite-testing: runs-on: ubuntu-18.04 if: "!contains(github.event.pull_request.labels.*.name, 'skip-ci')" steps: - name: checkout uses: actions/checkout@v2 with: fetch-depth: 0 - name: run RGW multisite test uses: ./.github/workflows/rgw-multisite-test with: github-token: ${{ secrets.GITHUB_TOKEN }} # ceph-image: # use default - name: upload test result uses: actions/upload-artifact@v2 if: always() with: name: rgw-multisite-testing path: test - name: setup tmate session for debugging when event is PR if: failure() && github.event_name == 'pull_request' uses: mxschmitt/action-tmate@v3 timeout-minutes: 60