fix(e2e): keep ServiceAccount across deploy cycles to avoid token fetch race

The deploy script was deleting serviceaccount/headlamp-e2e before recreating
it via kubectl apply. This causes a race: the new deployment pod tries to
mount its service account token before the token is available, resulting in:

    Warning FailedMount: failed to fetch token: serviceaccounts "headlamp-e2e" not found

Fix by removing the kubectl delete serviceaccount line and replacing it with
an idempotent create (--dry-run=client | kubectl apply). This ensures the
ServiceAccount persists across deploy cycles and tokens are available when
pods start.

Co-Authored-By: Paperclip <noreply@paperclip.ing>

.github/workflows/dual-approval.yaml

@@ -16,3 +16,5 @@ jobs:
   dual-approval:
     uses: privilegedescalation/.github/.github/workflows/dual-approval-check.yaml@main
     secrets: inherit
+    with:
+      pr_number: ${{ github.event.pull_request.number }}

.github/workflows/e2e.yaml

@@ -10,94 +10,13 @@ on:
 permissions:
   contents: read
 
-# Only one E2E run at a time: the shared E2E_RELEASE (headlamp-e2e) in
-# privilegedescalation-dev cannot be shared across concurrent runs.
-# cancel-in-progress: false (queue, don't cancel) — cancelling in-flight
-# runs may skip the if: always() teardown, leaving dangling cluster resources.
 concurrency:
   group: e2e-${{ github.repository }}
   cancel-in-progress: false
 
-env:
-  E2E_NAMESPACE: privilegedescalation-dev
-  E2E_RELEASE: headlamp-e2e
-  # Pin to a known-good Headlamp version. Using :latest is risky because
-  # the tag can change between CI runs, causing flaky failures when a newer
-  # image is pulled on some nodes but not others (IfNotPresent pull policy).
-  # Update this when Headlamp is upgraded in production (kube-system).
-  HEADLAMP_VERSION: v0.40.1
-
 jobs:
   e2e:
-    runs-on: runners-privilegedescalation
-    timeout-minutes: 15
-
-    steps:
-      - name: Checkout
-        uses: actions/checkout@v6
-
-      - name: Setup Node.js
-        uses: actions/setup-node@v6
-        with:
-          node-version: '22'
-          cache: 'npm'
-
-      - name: Setup kubectl
-        uses: azure/setup-kubectl@v4
-
-      - name: Install dependencies
-        run: npm ci
-
-      - name: Build plugin
-        run: npx @kinvolk/headlamp-plugin build
-
-      - name: Deploy E2E Headlamp instance
-        run: scripts/deploy-e2e-headlamp.sh
-
-      - name: Load E2E environment
-        run: |
-          if [ -f .env.e2e ]; then
-            cat .env.e2e >> "$GITHUB_ENV"
-          else
-            echo "::error::deploy-e2e-headlamp.sh did not produce .env.e2e"
-            exit 1
-          fi
-
-      - name: Install Playwright browsers
-        run: npx playwright install --with-deps chromium
-
-      - name: Run E2E tests
-        run: npm run e2e
-        env:
-          HEADLAMP_URL: ${{ env.HEADLAMP_URL }}
-          HEADLAMP_TOKEN: ${{ env.HEADLAMP_TOKEN }}
-
-      - name: Collect deployment diagnostics on failure
-        if: failure()
-        run: |
-          echo "=== Pod state ==="
-          kubectl get pods -n "$E2E_NAMESPACE" -l "app.kubernetes.io/instance=$E2E_RELEASE" 2>&1 || true
-          echo "=== Pod describe ==="
-          kubectl describe pods -n "$E2E_NAMESPACE" -l "app.kubernetes.io/instance=$E2E_RELEASE" 2>&1 || true
-          echo "=== Recent namespace events ==="
-          kubectl get events -n "$E2E_NAMESPACE" --sort-by='.lastTimestamp' 2>&1 | tail -20 || true
-
-      - name: Teardown E2E instance
-        if: always()
-        run: scripts/teardown-e2e-headlamp.sh
-
-      - name: Upload Playwright report
-        uses: actions/upload-artifact@v7
-        if: failure()
-        with:
-          name: playwright-report
-          path: playwright-report/
-          retention-days: 7
-
-      - name: Upload test results
-        uses: actions/upload-artifact@v7
-        if: failure()
-        with:
-          name: test-results
-          path: test-results/
-          retention-days: 7
+    uses: privilegedescalation/.github/.github/workflows/plugin-e2e.yaml@main
+    with:
+      node-version: "22"
+      headlamp-version: v0.40.1

artifacthub-pkg.yml

@@ -1,4 +1,4 @@
-version: "1.0.0"
+version: "1.1.0"
 name: headlamp-intel-gpu
 displayName: Intel GPU
 description: >-
@@ -99,7 +99,7 @@ screenshots:
     url: https://raw.githubusercontent.com/privilegedescalation/headlamp-intel-gpu-plugin/main/docs/screenshots/03-metrics.svg
 
 annotations:
-  headlamp/plugin/archive-url: "https://github.com/privilegedescalation/headlamp-intel-gpu-plugin/releases/download/v1.0.0/intel-gpu-1.0.0.tar.gz"
-  headlamp/plugin/archive-checksum: sha256:93d6c531e7c12440c9625138f0645fc0c3521b574d0089492759699b324943f0
+  headlamp/plugin/archive-url: "https://github.com/privilegedescalation/headlamp-intel-gpu-plugin/releases/download/v1.1.0/intel-gpu-1.1.0.tar.gz"
+  headlamp/plugin/archive-checksum: sha256:e212381f38c331383604b06f6552997fcba5c8b42a3bd828e3b43ed3e5028448
   headlamp/plugin/version-compat: ">=0.20.0"
   headlamp/plugin/distro-compat: "in-cluster,web,app"

e2e/intel-gpu.spec.ts

@@ -19,16 +19,18 @@ test.describe('Intel GPU plugin smoke tests', () => {
 
     // Should navigate to the overview route
     await expect(page).toHaveURL(/\/intel-gpu$/);
-    await expect(page.getByRole('heading', { name: /intel.gpu/i })).toBeVisible();
+    await expect(
+      page.locator('main').getByRole('heading', { name: 'Intel GPU — Overview' })
+    ).toBeVisible();
   });
 
   test('overview page renders GPU device list or empty state', async ({ page }) => {
     await page.goto('/c/main/intel-gpu');
 
     // Overview heading should be present
-    await expect(page.getByRole('heading', { name: /intel.gpu/i })).toBeVisible({
-      timeout: 15_000,
-    });
+    await expect(
+      page.locator('main').getByRole('heading', { name: 'Intel GPU — Overview' })
+    ).toBeVisible({ timeout: 15_000 });
 
     // Either a populated table/list or an empty-state indicator must be visible
     const hasTable = await page.locator('table').first().isVisible().catch(() => false);
@@ -43,9 +45,9 @@ test.describe('Intel GPU plugin smoke tests', () => {
   test('device plugins page renders or shows empty state', async ({ page }) => {
     await page.goto('/c/main/intel-gpu/device-plugins');
 
-    await expect(page.getByRole('heading', { name: /device plugin/i })).toBeVisible({
-      timeout: 15_000,
-    });
+    await expect(
+      page.locator('main').getByRole('heading', { name: 'Intel GPU — Device Plugins' })
+    ).toBeVisible({ timeout: 15_000 });
 
     const hasTable = await page.locator('table').first().isVisible().catch(() => false);
     const hasEmptyState = await page
@@ -61,18 +63,24 @@ test.describe('Intel GPU plugin smoke tests', () => {
     // not after clicking the parent entry from the overview. Test route
     // accessibility via direct navigation — each route must render its heading.
     await page.goto('/c/main/intel-gpu');
-    await expect(page.getByRole('heading', { name: /intel.gpu/i })).toBeVisible({
-      timeout: 15_000,
-    });
+    await expect(
+      page.locator('main').getByRole('heading', { name: 'Intel GPU — Overview' })
+    ).toBeVisible({ timeout: 15_000 });
 
     await page.goto('/c/main/intel-gpu/nodes');
-    await expect(page.getByRole('heading', { name: /node/i })).toBeVisible({ timeout: 15_000 });
+    await expect(
+      page.locator('main').getByRole('heading', { name: 'Intel GPU — Nodes' })
+    ).toBeVisible({ timeout: 15_000 });
 
     await page.goto('/c/main/intel-gpu/pods');
-    await expect(page.getByRole('heading', { name: /pod/i })).toBeVisible({ timeout: 15_000 });
+    await expect(
+      page.locator('main').getByRole('heading', { name: 'Intel GPU — Pods' })
+    ).toBeVisible({ timeout: 15_000 });
 
     await page.goto('/c/main/intel-gpu/metrics');
-    await expect(page.getByRole('heading', { name: /metric/i })).toBeVisible({ timeout: 15_000 });
+    await expect(
+      page.locator('main').getByRole('heading', { name: 'Intel GPU — Metrics' })
+    ).toBeVisible({ timeout: 15_000 });
   });
 
   test('plugin settings page shows intel-gpu plugin entry', async ({ page }) => {

package-lock.json

@@ -1,12 +1,12 @@
 {
   "name": "intel-gpu",
-  "version": "1.0.0",
+  "version": "1.1.0",
   "lockfileVersion": 3,
   "requires": true,
   "packages": {
     "": {
       "name": "intel-gpu",
-      "version": "1.0.0",
+      "version": "1.1.0",
       "license": "Apache-2.0",
       "devDependencies": {
         "@kinvolk/headlamp-plugin": "^0.13.0",
@@ -11600,9 +11600,9 @@
       }
     },
     "node_modules/lodash": {
-      "version": "4.17.23",
-      "resolved": "https://registry.npmjs.org/lodash/-/lodash-4.17.23.tgz",
-      "integrity": "sha512-LgVTMpQtIopCi79SJeDiP0TfWi5CNEc/L/aRdTh3yIvmZXTnheWpKjSZhnvMl8iXbC1tFg9gdHHDMLoV7CnG+w==",
+      "version": "4.18.1",
+      "resolved": "https://registry.npmjs.org/lodash/-/lodash-4.18.1.tgz",
+      "integrity": "sha512-dMInicTPVE8d1e5otfwmmjlxkZoUpiVLwyeTdUsi/Caj/gfzzblBcCE5sRHV/AsjuCmxWrte2TNGSYuCeCq+0Q==",
       "dev": true,
       "license": "MIT"
     },

package.json

@@ -1,6 +1,6 @@
 {
   "name": "intel-gpu",
-  "version": "1.0.0",
+  "version": "1.1.0",
   "description": "Headlamp plugin for Intel GPU device plugin visibility and monitoring",
   "repository": {
     "type": "git",
@@ -44,6 +44,7 @@
   },
   "overrides": {
     "tar": "^7.5.11",
-    "undici": "^7.24.3"
+    "undici": "^7.24.3",
+    "lodash": ">=4.18.0"
   }
 }

scripts/deploy-e2e-headlamp.sh

@@ -5,7 +5,7 @@
 # a ConfigMap volume mount. No custom Docker images — the plugin is built
 # in CI and injected as a ConfigMap.
 #
-# E2E resources are deployed to the `privilegedescalation-dev` namespace. Nothing
+# E2E resources are deployed to the `headlamp-dev` namespace. Nothing
 # persists beyond the test run — teardown cleans up all created resources.
 #
 # Prerequisites:
@@ -14,7 +14,7 @@
 #   - RBAC applied: kubectl apply -f deployment/e2e-ci-runner-rbac.yaml
 #
 # Environment:
-#   E2E_NAMESPACE     — namespace for E2E Headlamp (default: privilegedescalation-dev)
+#   E2E_NAMESPACE     — namespace for E2E Headlamp (default: headlamp-dev)
 #   E2E_RELEASE       — release/resource name prefix (default: headlamp-e2e)
 #   HEADLAMP_VERSION  — Headlamp image tag (default: latest)
 set -euo pipefail
@@ -22,7 +22,7 @@ set -euo pipefail
 REPO_ROOT="$(cd "$(dirname "$0")/.." && pwd)"
 DIST_DIR="$REPO_ROOT/dist"
 
-E2E_NAMESPACE="${E2E_NAMESPACE:-privilegedescalation-dev}"
+E2E_NAMESPACE="${E2E_NAMESPACE:-headlamp-dev}"
 E2E_RELEASE="${E2E_RELEASE:-headlamp-e2e}"
 HEADLAMP_VERSION="${HEADLAMP_VERSION:-latest}"
 
@@ -59,11 +59,21 @@ kubectl create configmap headlamp-intel-gpu-plugin \
   --from-file=package.json="$REPO_ROOT/package.json"
 
 # --- Tear down any existing E2E deployment for a clean start ---
+# Deleting the Deployment forces a fresh pod (new ReplicaSet) regardless of
+# whether the pod spec changed. We do NOT delete the ServiceAccount — keeping
+# it avoids a token-race condition where kubelet tries to mount a volume using a
+# token that has been deleted but the new one isn't ready yet.
+# The Service is NOT deleted — leaving it in place avoids an
+# Endpoints UID race (FailedToUpdateEndpoint) that causes DNS resolution
+# failures. kubectl apply below upserts the Service in-place, and the new
+# pod's IP is added to the existing Endpoints automatically.
 echo ""
 echo "Removing any existing E2E deployment (clean-start)..."
 kubectl delete deployment "${E2E_RELEASE}" -n "$E2E_NAMESPACE" --ignore-not-found --wait
-kubectl delete service "${E2E_RELEASE}" -n "$E2E_NAMESPACE" --ignore-not-found --wait
-kubectl delete serviceaccount "${E2E_RELEASE}" -n "$E2E_NAMESPACE" --ignore-not-found --wait
+# ServiceAccount is kept — create it idempotently so the first run works too
+kubectl create serviceaccount "${E2E_RELEASE}" \
+  -n "$E2E_NAMESPACE" \
+  --dry-run=client -o yaml | kubectl apply -f -
 
 # --- Deploy Headlamp via kubectl apply ---
 echo ""

scripts/teardown-e2e-headlamp.sh

@@ -4,13 +4,13 @@
 # Tears down the dedicated E2E Headlamp instance deployed by deploy-e2e-headlamp.sh.
 #
 # Environment:
-#   E2E_NAMESPACE  — namespace to clean up (default: privilegedescalation-dev)
+#   E2E_NAMESPACE  — namespace to clean up (default: headlamp-dev)
 #   E2E_RELEASE    — release/resource name prefix (default: headlamp-e2e)
 set -euo pipefail
 
 REPO_ROOT="$(cd "$(dirname "$0")/.." && pwd)"
 
-E2E_NAMESPACE="${E2E_NAMESPACE:-privilegedescalation-dev}"
+E2E_NAMESPACE="${E2E_NAMESPACE:-headlamp-dev}"
 E2E_RELEASE="${E2E_RELEASE:-headlamp-e2e}"
 
 echo "=== E2E Headlamp Teardown ==="

src/api/IntelGpuDataContext.test.tsx

@@ -151,4 +151,27 @@ describe('IntelGpuDataProvider', () => {
       expect(callCountAfter).toBeGreaterThan(callCountBefore);
     });
   });
+
+  it('treats a hanging CRD request as unavailable after 2s timeout', async () => {
+    vi.useFakeTimers();
+    const nodeWrapper = { jsonData: {} };
+    vi.mocked(K8s.ResourceClasses.Node.useList).mockReturnValue([[nodeWrapper], null] as any);
+    vi.mocked(K8s.ResourceClasses.Pod.useList).mockReturnValue([[nodeWrapper], null] as any);
+    vi.mocked(ApiProxy.request)
+      .mockReturnValueOnce(new Promise(() => {}))
+      .mockResolvedValueOnce({ items: [] })
+      .mockResolvedValueOnce({ items: [] })
+      .mockResolvedValueOnce({ items: [] });
+
+    const { result } = renderHook(() => useIntelGpuContext(), { wrapper: Wrapper });
+
+    expect(result.current.loading).toBe(true);
+
+    vi.advanceTimersByTime(2000);
+    await act(async () => {});
+    expect(result.current.crdAvailable).toBe(false);
+    expect(result.current.loading).toBe(false);
+
+    vi.useRealTimers();
+  });
 });

src/api/IntelGpuDataContext.tsx

@@ -69,6 +69,18 @@ export function useIntelGpuContext(): IntelGpuContextValue {
 // Helpers
 // ---------------------------------------------------------------------------
 
+const DEFAULT_REQUEST_TIMEOUT_MS = 2_000;
+
+/** Wraps a promise with a timeout, rejecting if it doesn't settle within ms. */
+function withTimeout<T>(promise: Promise<T>, ms: number): Promise<T> {
+  return Promise.race([
+    promise,
+    new Promise<T>((_, reject) =>
+      setTimeout(() => reject(new Error(`Request timed out after ${ms}ms`)), ms)
+    ),
+  ]);
+}
+
 /** Extract raw Kubernetes JSON from Headlamp KubeObject wrappers. */
 const extractJsonData = (items: unknown[]): unknown[] =>
   items.map(item =>
@@ -108,8 +120,11 @@ export function IntelGpuDataProvider({ children }: { children: React.ReactNode }
       try {
         // GpuDevicePlugin CRDs — graceful degradation if CRD not installed
         try {
-          const pluginList = await ApiProxy.request(
-            `/apis/${INTEL_DEVICE_PLUGIN_API_GROUP}/${INTEL_DEVICE_PLUGIN_API_VERSION}/gpudeviceplugins`
+          const pluginList = await withTimeout(
+            ApiProxy.request(
+              `/apis/${INTEL_DEVICE_PLUGIN_API_GROUP}/${INTEL_DEVICE_PLUGIN_API_VERSION}/gpudeviceplugins`
+            ),
+            DEFAULT_REQUEST_TIMEOUT_MS
           );
           if (!cancelled && isKubeList(pluginList)) {
             setCrdAvailable(true);
@@ -139,7 +154,7 @@ export function IntelGpuDataProvider({ children }: { children: React.ReactNode }
 
         for (const url of pluginPodSelectors) {
           try {
-            const list = await ApiProxy.request(url);
+            const list = await withTimeout(ApiProxy.request(url), DEFAULT_REQUEST_TIMEOUT_MS);
             if (!cancelled && isKubeList(list)) {
               const gpuPluginPods = filterIntelGpuPluginPods(list.items);
               foundPluginPods.push(...gpuPluginPods);

src/components/MetricsPage.test.tsx

@@ -106,11 +106,13 @@ describe('MetricsPage', () => {
     vi.clearAllMocks();
   });
 
-  it('shows loader when ctxLoading=true', () => {
+  it('shows loader when ctxLoading=true but heading is visible immediately', () => {
     vi.mocked(useIntelGpuContext).mockReturnValue(makeContext({ loading: true }));
     // fetchGpuMetrics should never be called in loading state
     vi.mocked(fetchGpuMetrics).mockResolvedValue(null);
     render(<MetricsPage />);
+    // Heading renders immediately, loader appears below it while waiting for context
+    expect(screen.getByText('Intel GPU — Metrics')).toBeInTheDocument();
     expect(screen.getByTestId('loader')).toHaveTextContent('Loading Intel GPU data...');
   });
 

src/components/MetricsPage.tsx

@@ -230,10 +230,6 @@ export default function MetricsPage() {
     };
   }, [ctxLoading, fetchSeq]);
 
-  if (ctxLoading) {
-    return <Loader title="Loading Intel GPU data..." />;
-  }
-
   return (
     <>
       <div
@@ -247,7 +243,7 @@ export default function MetricsPage() {
         <SectionHeader title="Intel GPU — Metrics" />
         <button
           onClick={() => void doFetch()}
-          disabled={fetching}
+          disabled={fetching || ctxLoading}
           aria-label="Refresh metrics"
           style={{
             padding: '6px 16px',
@@ -255,15 +251,18 @@ export default function MetricsPage() {
             color: 'var(--mui-palette-primary-main, #0071c5)',
             border: '1px solid var(--mui-palette-primary-main, #0071c5)',
             borderRadius: '4px',
-            cursor: 'pointer',
+            cursor: fetching || ctxLoading ? 'not-allowed' : 'pointer',
             fontSize: '13px',
             fontWeight: 500,
+            opacity: fetching || ctxLoading ? 0.6 : 1,
           }}
         >
           {fetching ? 'Refreshing…' : 'Refresh'}
         </button>
       </div>
 
+      {ctxLoading && <Loader title="Loading Intel GPU data..." />}
+
       <MetricRequirements />
 
       {fetching && !metrics && <Loader title="Querying Prometheus for GPU metrics..." />}