ci/cli: add full backup/restore test

This test simulates migration of a Rancher Manager server.

Signed-off-by: Loic Devulder <[email protected]>

.github/workflows/SCHEDULING.md

@@ -14,12 +14,13 @@ We try to spread the tests as best as we can to avoid SPOT issue as well as not
 | CLI Multicluster | Sunday | 5am | us-central1-b |
 | CLI Regression | Saturday | 11am | us-central1-c |
 | CLI Rancher Manager Devel | Sunday | 8am | us-central1-c |
-| UI Rancher Manager Devel | Sunday | 12pm | us-central1-a |
 | CLI K3s Downgrade | Sunday | 2pm | us-central1-b |
+| CLI Full backup/restore (migration) | Sunday | 4pm | us-central1-c |
 | UI K3s | Monday to Saturday | 2am | us-central1-a |
 | UI K3s Upgrade | Monday to Saturday | 4am | us-central1-a |
 | UI RKE2 | Monday to Saturday | 2am | us-central1-b |
 | UI RKE2 Upgrade | Monday to Saturday | 4am | us-central1-b |
+| UI Rancher Manager Devel | Sunday | 12pm | us-central1-a |
 | Update tests description | All days | 11pm | us-central1 |
 
 **NOTE:** please note that the GitHub scheduler uses UTC and our GCP runners are deployed in `us-central1`, so UTC-5.

.github/workflows/cli-full-backup-restore-matrix.yaml

@@ -0,0 +1,53 @@
+# This workflow calls the master E2E workflow with custom variables
+name: CLI-Full-Backup-Restore
+
+on:
+  workflow_dispatch:
+    inputs:
+      destroy_runner:
+        description: Destroy the auto-generated self-hosted runner
+        default: true
+        type: boolean
+      k8s_downstream_version:
+        description: Rancher cluster downstream version to use
+        default: '"v1.27.13+k3s1"'
+        type: string
+      k8s_upstream_version:
+        description: Rancher cluster upstream version to use
+        default: '"v1.27.13+k3s1"'
+        type: string
+      qase_run_id:
+        description: Qase run ID where the results will be reported
+        type: string
+      rancher_version:
+        description: Rancher Manager channel/version/head_version to use
+        default: '"stable/latest"'
+        type: string
+  schedule:
+    # Every Sunday at 4pm UTC (11am in us-central1)
+    - cron: '0 16 * * 0'
+
+jobs:
+  cli:
+    strategy:
+      fail-fast: false
+      max-parallel: 4
+      matrix:
+        k8s_downstream_version: ${{ fromJSON(format('[{0}]', inputs.k8s_downstream_version || '"v1.30.4+k3s1","v1.30.4+rke2r1"')) }}
+        k8s_upstream_version: ${{ fromJSON(format('[{0}]', inputs.k8s_upstream_version || '"v1.30.4+k3s1","v1.30.4+rke2r1"')) }}
+        rancher_version: ${{ fromJSON(format('[{0}]', inputs.rancher_version || '"latest/devel/2.9"')) }}
+    uses: ./.github/workflows/master_e2e.yaml
+    secrets:
+      credentials: ${{ secrets.GCP_CREDENTIALS }}
+      pat_token: ${{ secrets.SELF_HOSTED_RUNNER_PAT_TOKEN }}
+      qase_api_token: ${{ secrets.QASE_API_TOKEN }}
+    with:
+      destroy_runner: ${{ github.event_name == 'schedule' && true || inputs.destroy_runner }}
+      full_backup_restore: true
+      k8s_downstream_version: ${{ matrix.k8s_downstream_version }}
+      k8s_upstream_version: ${{ matrix.k8s_upstream_version }}
+      node_number: 3
+      qase_run_id: ${{ github.event_name == 'schedule' && 'auto' || inputs.qase_run_id }}
+      rancher_version: ${{ matrix.rancher_version }}
+      test_type: cli
+      zone: us-central1-c

.github/workflows/master_e2e.yaml

@@ -45,6 +45,10 @@ on:
         description: Force OS downgrade
         default: false
         type: boolean
+      full_backup_restore:
+         description: Test migration-like backup/restore functionality
+         default: false
+         type: boolean
       k8s_downstream_version:
         description: Rancher cluster downstream version to use
         type: string
@@ -228,6 +232,7 @@ jobs:
       destroy_runner: ${{ inputs.destroy_runner }}
       elemental_ui_version: ${{ inputs.elemental_ui_version }}
       force_downgrade: ${{ inputs.force_downgrade }}
+      full_backup_restore: ${{ inputs.full_backup_restore }}
       k8s_downstream_version: ${{ inputs.k8s_downstream_version }}
       node_number: ${{ inputs.node_number }}
       operator_repo: ${{ inputs.operator_repo }}

.github/workflows/sub_cli.yaml

@@ -32,6 +32,9 @@ on:
       force_downgrade:
         required: true
         type: boolean
+      full_backup_restore:
+        required: true
+        type: boolean
       k8s_downstream_version:
         required: true
         type: string
@@ -350,10 +353,25 @@ jobs:
         run: |
           cd tests && make e2e-upgrade-node && make e2e-check-app
 
-      - name: Test Backup/Restore Elemental resources with Rancher Manager
+      - name: Test Backup/Restore Rancher Manager/Elemental resources
         id: test_backup_restore
+        env:
+          CA_TYPE: ${{ inputs.ca_type }}
+          OPERATOR_REPO: ${{ inputs.operator_repo }}
+          PUBLIC_FQDN: ${{ inputs.public_fqdn }}
+          PUBLIC_DOMAIN: ${{ inputs.public_domain }}
         run: |
-          cd tests && make e2e-backup-restore && make e2e-check-app
+          cd tests
+
+          # Run simple or full backup/restore test
+          if ${{ inputs.full_backup_restore == true }}; then
+            make e2e-full-backup-restore
+          else
+            make e2e-simple-backup-restore
+          fi
+
+          # Check the installed application
+          make e2e-check-app
 
       - name: Extract ISO version
         id: iso_version
@@ -426,8 +444,8 @@ jobs:
         env:
           OPERATOR_REPO: ${{ inputs.operator_repo }}
         # Don't test Operator uninstall if we want to keep the runner for debugging purposes
-        if: ${{ inputs.destroy_runner == true }}
-        run: cd tests && make e2e-uninstall-operator
+        if: ${{ inputs.destroy_runner == true && inputs.full_backup_restore == false }}
+        run: cd tests && make e2e-uninstall-operator && make e2e-check-app
 
       # This step must be called in each worklow that wants a summary!
       - name: Get logs and add summary

.github/workflows/sub_test_choice.yaml

@@ -41,6 +41,9 @@ on:
       force_downgrade:
         required: true
         type: boolean
+      full_backup_restore:
+        required: true
+        type: boolean
       k8s_downstream_version:
         required: true
         type: string
@@ -171,6 +174,7 @@ jobs:
       cluster_type: ${{ inputs.cluster_type }}
       destroy_runner: ${{ inputs.destroy_runner }}
       force_downgrade: ${{ inputs.force_downgrade }}
+      full_backup_restore: ${{ inputs.full_backup_restore }}
       k8s_downstream_version: ${{ inputs.k8s_downstream_version }}
       node_number: ${{ inputs.node_number }}
       operator_repo: ${{ inputs.operator_repo }}

README.md

@@ -20,6 +20,7 @@
 [![CLI-K3s-Downgrade](https://github.com/rancher/elemental/actions/workflows/cli-k3s-downgrade-matrix.yaml/badge.svg)](https://github.com/rancher/elemental/actions/workflows/cli-k3s-downgrade-matrix.yaml)
 [![CLI-K3s-Scalability](https://github.com/rancher/elemental/actions/workflows/cli-k3s-scalability-matrix.yaml/badge.svg)](https://github.com/rancher/elemental/actions/workflows/cli-k3s-scalability-matrix.yaml)
 [![CLI-K3s-SELinux](https://github.com/rancher/elemental/actions/workflows/cli-k3s-selinux-matrix.yaml/badge.svg)](https://github.com/rancher/elemental/actions/workflows/cli-k3s-selinux-matrix.yaml)
+[![CLI-Full-Backup-Restore](https://github.com/rancher/elemental/actions/workflows/cli-full-backup-restore-matrix.yaml/badge.svg)](https://github.com/rancher/elemental/actions/workflows/cli-full-backup-restore-matrix.yaml)
 [![CLI-Multicluster](https://github.com/rancher/elemental/actions/workflows/cli-multicluster-matrix.yaml/badge.svg)](https://github.com/rancher/elemental/actions/workflows/cli-multicluster-matrix.yaml)
 [![CLI-Rancher-Manager-Head-2.7](https://github.com/rancher/elemental/actions/workflows/cli-rm-head-2.7-matrix.yaml/badge.svg)](https://github.com/rancher/elemental/actions/workflows/cli-rm-head-2.7-matrix.yaml)
 [![CLI-Regression](https://github.com/rancher/elemental/actions/workflows/cli-regression-matrix.yaml/badge.svg)](https://github.com/rancher/elemental/actions/workflows/cli-regression-matrix.yaml)

tests/Makefile

@@ -61,15 +61,15 @@ e2e-airgap-rancher: deps
 e2e-bootstrap-node: deps
 	ginkgo --timeout $(GINKGO_TIMEOUT)s --label-filter bootstrap -r -v ./e2e
 
-e2e-backup-restore: deps
-	ginkgo --label-filter test-backup-restore -r -v ./e2e
-
 e2e-check-app: deps
 	ginkgo --label-filter check-app -r -v ./e2e
 
 e2e-configure-rancher: deps
 	ginkgo --label-filter configure -r -v ./e2e
 
+e2e-full-backup-restore: deps
+	ginkgo --label-filter test-full-backup-restore -r -v ./e2e
+
 e2e-get-logs: deps
 	ginkgo --label-filter logs -r -v ./e2e
 
@@ -91,11 +91,14 @@ e2e-iso-image: deps
 e2e-multi-cluster: deps
 	ginkgo --timeout $(GINKGO_TIMEOUT)s --label-filter multi-cluster -r -v ./e2e
 
+e2e-prepare-archive: deps
+	ginkgo --label-filter prepare-archive -r -v ./e2e
+
 e2e-reset: deps
 	ginkgo --label-filter reset -r -v ./e2e
 
-e2e-prepare-archive: deps
-	ginkgo --label-filter prepare-archive -r -v ./e2e
+e2e-simple-backup-restore: deps
+	ginkgo --label-filter test-simple-backup-restore -r -v ./e2e
 
 e2e-ui-rancher: deps
 	ginkgo --label-filter ui -r -v ./e2e

tests/assets/restore.yaml

@@ -7,4 +7,4 @@ metadata:
 spec:
   backupFilename: %BACKUP_FILE%
   deleteTimeoutSeconds: 10
-  prune: true
+  prune: %PRUNE%