Add testing DSS with Canonical Kubernetes (New)

contrib/checkbox-dss-validation/README.md

@@ -21,7 +21,7 @@ lxd init --auto
 git clone https://github.com/canonical/checkbox
 cd checkbox/contrib/checkbox-dss-validation
 snapcraft
-sudo snap install --dangerous --classic ./checkbox-dss_3.0_amd64.snap
+sudo snap install --dangerous --classic ./checkbox-dss_3.1_amd64.snap
 ```
 
 Make sure that the provider service is running and active:
@@ -38,21 +38,37 @@ Some test need dependencies, and a helper script is available to install them:
 checkbox-dss.install-deps
 ```
 
-By default this will install the `data-science-stack` snap from the `latest/stable`
-channel. To instead install from `latest/edge` use:
+By default this will install the `data-science-stack` snap from the `1.0/stable`
+channel. To instead install from `1.1/edge` use:
 
 ```shell
-checkbox-dss.install-deps --dss-snap-channel latest/edge
+checkbox-dss.install-deps --dss-snap-channel 1.1/edge
 ```
 
-Furthermore, the default `microk8s` snap channel is `1.28/stable` in classic mode,
-but this can be customized as
+Furthermore, by default, `microk8s` snap from channel `1.28/stable` is installed
+in classic mode.
+The channel for `microk8s` can be customized as
 (please note that this snap must to be `--classic` to enable GPU support):
 
 ```shell
-checkbox-dss.install-deps --microk8s-snap-channel 1.31/stable
+checkbox-dss.install-deps --microk8s-snap-channel 1.32/stable
 ```
 
+It is also possible to install [Canonical Kubernetes](https://snapcraft.io/k8s)
+**instead** of `microk8s`.
+**Only one of, and at least one of either**
+`microk8s` or Canonical Kubernetes must be installed.
+Canonical Kubernetes will be installed instead of `microk8s` if the argument
+`--canonical-k8s-snap-channel` is provided with a value:
+
+```shell
+checkbox-dss.install-deps --canonical-k8s-snap-channel 1.32/stable --dss-snap-channel 1.1/edge
+```
+
+When Canonical Kubernetes is install, the snape `helm` will also be installed from the
+default channel in `--classic` mode.
+`helm` will be used to enable NVIDIA GPU in the Canonical Kubernetes cluster.
+
 These validations also need the `kubectl` snap installed, and the default channel
 used for that is `1.29/stable`, but can be customized as shown previously by passing
 the appropriate channel name for `--kubectl-snap-channel`.

contrib/checkbox-dss-validation/bin/install-deps

@@ -1,7 +1,7 @@
 #!/bin/bash
 set -e
 
-dss_snap_channel="latest/stable"
+dss_snap_channel="1.0/stable"
 microk8s_snap_channel="1.28/stable"
 kubectl_snap_channel="1.29/stable"
 
@@ -35,21 +35,41 @@ setup_microk8s_snap() {
         deployment/coredns \
         deployment/hostpath-provisioner
     sudo microk8s.kubectl -n kube-system rollout status ds/calico-node
+
+    # hack as redirecting stdout anywhere but /dev/null throws a permission denied error
+    # see: https://forum.snapcraft.io/t/eksctl-cannot-write-to-stdout/17254/4
+    sudo microk8s.kubectl config view --raw | tee "${SNAP_REAL_HOME}/.kube/config" >/dev/null
 }
 
 setup_kubectl_snap() {
     # This is needed to overcome the following bug within microk8s:
     # https://github.com/canonical/microk8s/issues/4453
     echo -e "\nInstalling kubectl snap from channel $1"
     sudo snap install kubectl --classic --channel="$1"
+}
+
+setup_canonical_k8s_snap() {
+    echo -e "\nInstalling Canonical k8s snap from channel $1"
+    sudo snap install k8s --classic --channel "$1"
+    sudo k8s bootstrap
+    sudo k8s enable local-storage
+
+    # Directory needed for kubeconfig
+    mkdir -p "${SNAP_REAL_HOME}/.kube"
+
     # hack as redirecting stdout anywhere but /dev/null throws a permission denied error
     # see: https://forum.snapcraft.io/t/eksctl-cannot-write-to-stdout/17254/4
-    sudo microk8s.kubectl config view --raw | tee "${SNAP_REAL_HOME}/.kube/config" >/dev/null
+    sudo k8s config | tee "${SNAP_REAL_HOME}/.kube/config" >/dev/null
 }
 
 help_function() {
-    echo "This script is used install all dependencies for checkbox-dss to run; defaults for optional arguments are shown in usage"
-    echo "Usage: checkbox-dss.install-deps [--dss-snap-channel $dss_snap_channel] [--microk8s-snap-channel $microk8s_snap_channel] [--kubectl-snap-channel $kubectl_snap_channel]"
+    echo "This script is used install all dependencies for checkbox-dss to run;"
+    echo "defaults for optional arguments are shown in usage except for"
+    echo "'--canonical-k8s-channel':"
+    echo "if this arg is not provided, then microk8s will be installed from its default value,"
+    echo "otherwise Canonical k8s will be installed from the given channel instead of microk8s."
+    echo "Usage: checkbox-dss.install-deps [--dss-snap-channel $dss_snap_channel] [--canonical-k8s-channel NO-DEFAULT | --microk8s-snap-channel $microk8s_snap_channel] [--kubectl-snap-channel $kubectl_snap_channel]"
+    exit 2
 }
 
 main() {
@@ -67,12 +87,23 @@ main() {
             kubectl_snap_channel="$2"
             shift 2
             ;;
+        --canonical-k8s-snap-channel)
+            canonical_k8s_snap_channel="$2"
+            shift 2
+            ;;
         *) help_function ;;
         esac
     done
 
-    echo -e "\n Step 1/4: Setting up microk8s"
-    setup_microk8s_snap "$microk8s_snap_channel"
+    if [[ -n "$canonical_k8s_snap_channel" ]]; then
+        echo -e "\n Step 1/4: Setting up Canonical k8s and helm"
+        setup_canonical_k8s_snap "$canonical_k8s_snap_channel"
+    else
+        echo -e "\n Step 1/4: Setting up microk8s and helm"
+        setup_microk8s_snap "$microk8s_snap_channel"
+    fi
+    echo -e "\n Installing the helm snap"
+    sudo snap install helm --classic
 
     echo -e "\n Step 2/4: Setting up kubectl"
     setup_kubectl_snap "$kubectl_snap_channel"

contrib/checkbox-dss-validation/checkbox-provider-dss/bin/check_nvidia_gpu_rollout.sh

@@ -21,7 +21,7 @@ set -eou pipefail
 
 NAMESPACE="${1:-"gpu-operator-resources"}"
 sleep 10
-kubectl -n "$NAMESPACE" rollout status ds/gpu-operator-node-feature-discovery-worker
+kubectl -n "$NAMESPACE" rollout status ds/gpu-operator-feature-discovery
 sleep 10
 kubectl -n "$NAMESPACE" rollout status ds/nvidia-device-plugin-daemonset
 sleep 10

contrib/checkbox-dss-validation/checkbox-provider-dss/units/jobs.pxu

@@ -4,7 +4,6 @@ flags: simple
 imports: from com.canonical.certification import executable
 requires:
   executable.name == 'dss'
-  executable.name == 'microk8s'
 _summary: Check that the DSS environment initializes
 estimated_duration: 2m
 command:
@@ -254,31 +253,37 @@ command: run_dss.sh remove pytorch-intel
 
 # NVIDIA CUDA jobs ############################################################
 
-id: microk8s_nvidia_gpu_addon/enable
+id: nvidia_gpu_operator/install
 category_id: dss-regress
 flags: simple
 imports:
   from com.canonical.certification import executable
   from com.canonical.certification import graphics_card
 requires:
   graphics_card.vendor == 'NVIDIA Corporation'
-  executable.name == 'microk8s'
+  executable.name == 'helm'
   executable.name == 'kubectl'
 depends: dss/initialize
-_summary: Enable NVIDIA GPU addon in microk8s
+_summary: Install NVIDIA GPU operator in the Kubernetes cluster
 estimated_duration: 10m
 command:
   set -eou pipefail
-  OPERATOR_VERSION="24.6.2"
-  microk8s enable gpu --driver=operator --version="${OPERATOR_VERSION}"
-  check_cuda_rollout.sh
+  helm repo add nvidia https://helm.ngc.nvidia.com/nvidia
+  helm repo update
+  helm install --wait --generate-name -n gpu-operator-resources --create-namespace nvidia/gpu-operator --kubeconfig ~/.kube/config
+  sleep 15
+  check_nvidia_gpu_rollout.sh
 
-id: nvidia_gpu_addon/validations_succeed
+id: nvidia_gpu_operator/validations_succeed
 category_id: dss-regress
 flags: simple
-imports: from com.canonical.certification import executable
-requires: executable.name == 'kubectl'
-depends: microk8s_nvidia_gpu_addon/enable
+imports:
+  from com.canonical.certification import executable
+  from com.canonical.certification import graphics_card
+requires:
+  graphics_card.vendor == 'NVIDIA Corporation'
+  executable.name == 'kubectl'
+depends: nvidia_gpu_operator/install
 _summary: NVIDIA GPU validations should succeed
 estimated_duration: 10s
 command:
@@ -292,7 +297,7 @@ category_id: dss-regress
 flags: simple
 imports: from com.canonical.certification import executable
 requires: executable.name == 'dss'
-depends: nvidia_gpu_addon/validations_succeed
+depends: nvidia_gpu_operator/validations_succeed
 _summary: Check that dss status reports that NVIDIA GPU acceleration is enabled
 estimated_duration: 5s
 command:

contrib/checkbox-dss-validation/checkbox-provider-dss/units/test-plan.pxu

@@ -26,8 +26,8 @@ include:
     dss/create_pytorch_intel_notebook
     xpu/pytorch_can_use_xpu
     dss/remove_pytorch_intel_notebook
-    microk8s_nvidia_gpu_addon/enable
-    nvidia_gpu_addon/validations_succeed
+    nvidia_gpu_operator/install
+    nvidia_gpu_operator/validations_succeed
     dss/status_nvidia_gpu
     dss/create_pytorch_cuda_notebook
     cuda/pytorch_can_use_cuda

contrib/checkbox-dss-validation/snap/snapcraft.yaml

@@ -2,7 +2,7 @@ name: checkbox-dss
 summary: Checkbox tests for validating the Data Science Stack
 description: |
   Collection of tests to be run on devices that are part of the dss-validation project
-version: '3.0'
+version: '3.1'
 confinement: classic
 grade: stable