The heading 'Intel GPU — Metrics' was blocked behind the ctxLoading check,
causing the E2E navigation test to timeout when navigating directly to
/c/main/intel-gpu/metrics. The K8s.ResourceClasses.useList() hooks
in IntelGpuDataContext can take time to resolve when navigating directly
to the metrics route (as opposed to via sidebar), causing ctxLoading to
remain true beyond the 15s test timeout.
Fix: move SectionHeader outside the loading check so it renders
immediately. The Loader now appears below the heading while waiting
for context to load. Also disable the Refresh button during ctxLoading.
Updated unit test to verify heading is visible even when ctxLoading=true.
Fixes: headlamp-intel-gpu-plugin#42
Co-Authored-By: Paperclip <noreply@paperclip.ing>
* test: add component test coverage for all untested files
Adds 60 new tests (108 total) covering every untested module:
- IntelGpuDataContext: provider renders, loading/loaded states, CRD
available/unavailable paths, refresh, useIntelGpuContext throws outside
provider
- OverviewPage: loading, plugin-not-detected, error, populated, refresh
button, CRD notice, device plugin table, plugin daemon pods, active pods
- NodesPage: loading, empty state, GPU node summary table, detail cards
- PodsPage: loading, empty state, summary counts, pending pod attention,
all-pods table
- DevicePluginsPage: loading, CRD unavailable, no-plugins, plugin detail,
daemon pod table
- NodeDetailSection: null for non-GPU nodes, GPU capacity/allocatable rows,
pod list, loading state
- PodDetailSection: null for non-GPU pods, GPU resource rows, phase status,
limits-only containers
- MetricsPage: context loading gate, Prometheus unreachable, empty chips,
chip cards with power values, MetricRequirements always rendered, refresh
Also fixes vitest.config.mts to pin NODE_ENV=test so tests run correctly
without requiring callers to set it explicitly.
Co-Authored-By: Paperclip <noreply@paperclip.ing>
* fix: remove unused act import and merge duplicate metrics imports in MetricsPage.test.tsx
Co-Authored-By: Paperclip <noreply@paperclip.ing>
* fix: cast useList mock return values to any in IntelGpuDataContext.test.tsx
The Headlamp useList() return type is an intersection of a tuple and
QueryListResponse, which plain array literals like [[], null] and
[null, null] do not satisfy. Cast all useList mockReturnValue arguments
to any so tsc passes without requiring full KubeObject stub objects.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
* style: run Prettier formatting and ESLint lint:fix on test files
Addresses CI format:check failures and import-sort warning in
MetricsPage.test.tsx flagged by QA on PR #17.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
---------
Co-authored-by: Hugh Hackman <hugh@privilegedescalation.com>
Co-authored-by: Paperclip <noreply@paperclip.ing>
Co-authored-by: Gandalf the Greybeard <gandalf@privilegedescalation.com>
Co-authored-by: Gandalf the Greybeard <gandalf@privilegedescalation.dev>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
Co-authored-by: Gandalf the Greybeard <gandalf-the-greybeard[bot]@users.noreply.github.com>
Artifact Hub listing was renamed with new repository ID
3c97f78a-26e3-4e8a-89e7-29884602e3d7. Updates package name,
sidebar entries, routes, archive URL, and documentation.
Refs: PRI-26
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Remove unsafe `as any` casts, fix MetricsPage fetch cancellation safety,
delete dead AppBarGpuBadge component, fix typo in data context, move
extractJsonData to module scope, resolve ESLint/Prettier indent conflict,
fix artifacthub-pkg.yml version mismatch and inaccurate description.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Auto-fix import ordering, quote style, and indentation via eslint --fix
and prettier --write. Remove unused variable in NodesPage and PodsPage.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add a file-level comment and in-page requirements section explaining
exactly what is and isn't available for each metric type:
Power (W) -- available on discrete GPU nodes via node-exporter
hwmon collector + i915 driver (no extra config)
Frequency (MHz) -- NOT available; node-exporter --collector.drm is
AMD-only and does not read i915 gt_freq sysfs
Utilization (%) -- NOT available; no standard Prometheus collector
supports i915 engine busy metrics
iGPU nodes -- no metrics at all (iGPU driver has no hwmon)
The in-page MetricRequirements component surfaces this information
directly in the UI so operators know what to expect and why.
Generated with [Claude Code](https://claude.ai/code)
via [Happy](https://happy.engineering)
Co-Authored-By: Claude <noreply@anthropic.com>
Co-Authored-By: Happy <yesreply@happy.engineering>
The Intel GPU device plugin -enable-monitoring flag registers a monitoring
K8s resource type (not a Prometheus endpoint). Real GPU power metrics come
from node-exporter's hwmon collector which scrapes the i915 kernel driver.
- Rewrite src/api/metrics.ts: query kube-prometheus-stack Prometheus for
node_hwmon_energy_joule_total (rate → watts), node_hwmon_power_max_watt
(TDP), joined with node_hwmon_chip_names{chip_name="i915"} to identify
GPU chips. Instance → node name resolved via node_uname_info.
- Rewrite src/components/MetricsPage.tsx: shows per-chip current power (W)
with bar vs TDP, total fleet power summary, last-fetched timestamp.
Auto-discovers Prometheus service in monitoring namespace.
- Update artifacthub-pkg.yml checksum for repackaged v0.2.0 tarball.
Generated with [Claude Code](https://claude.ai/code)
via [Happy](https://happy.engineering)
Co-Authored-By: Claude <noreply@anthropic.com>
Co-Authored-By: Happy <yesreply@happy.engineering>
privilegedescalation-engineer
Chris Farhood