- Accelerator images (cuda, gpu, gpu-legacy, rocm) no longer report OK when their GPU or NPU is not working at startup: the startup log shows FAIL,
/healthreportsok: false, every page shows a red line in the header, and the System page gives cause and fix under the Backend tile. - cuda: the startup log now names the exact case and lists what arrives in the container.
- Docs: the Compose example for NVIDIA now names only
capabilities: [gpu]. Withvideoadded,nvidia-smiwas missing in the container; if you copied the old example, removevideo. - System page: the Analysis tile turns red when the last analysis has failed.
gpu-legacy: not load-tested before this release (no test machine for this variant).
rocm: not load-tested before this release (no test machine for this variant); a field installation last confirmed this variant on 0.1.0.511.
Known issues, accepted for this release:
- cuda: in the second load test the warning counter of the service in /health was one line behind the log at the moment of the last snapshot (62 against 63); the log itself was complete, and every event was booked once with the persons of the reference run.