KernelCI Dashboard Report - December 08 - December 19, 2025

Hello everyone,

Over the last two weeks we focused on improving ingester observability
and reliability, extending metrics, and advancing the hardware view
improvements, including starting the rollout of a significantly faster
/hardware view (currently staging-only at
https://staging.dashboard.kernelci.org:9000/hardware-new), no existing
endpoints/routes changed, we only added a new backend endpoint and a
new frontend route, feel free to test it out!

This will be the last report of the year, as we=E2=80=99ll be taking a brea=
k
for Christmas and New Year=E2=80=99s Eve. We=E2=80=99ll resume reports in t=
he new
year.

## Ingester

- Added submissions and items-created Prometheus metrics. [1]
- Switched the ingester execution model from threads to processes. [2]
- Reported ingester queue size via OpenMetrics. [3]
- Added a `pending_retry` folder to improve handling of retryable
ingestion failures. [4]
- Fixed log excerpt handling by ensuring `log_excerpt` is set to an
empty string after conversion. [5]

## UI

- Added a lab filter to the hardware details view. [6]
- Added a new hardware listing view backed by the hardware-v2 endpoint. [7]

## Notifications

- Added a metrics summary notification and template. [8]
- Fixed the metrics notification recipient configuration. [9]

## Database

- Updated the latest_checkout table primary key via migration. [10]

## References

[1] - https://github.com/kernelci/dashboard/pull/1660
[2] - https://github.com/kernelci/dashboard/pull/1670
[3] - https://github.com/kernelci/dashboard/pull/1677
[4] - https://github.com/kernelci/dashboard/pull/1678
[5] - https://github.com/kernelci/dashboard/pull/1682
[6] - https://github.com/kernelci/dashboard/pull/1671
[7] - https://github.com/kernelci/dashboard/pull/1685
[8] - https://github.com/kernelci/dashboard/pull/1646
[9] - https://github.com/kernelci/dashboard/pull/1683
[10] - https://github.com/kernelci/dashboard/pull/1681

Merry christmas and happy new year!