Hi,
> > Alert fatigue is a well known phenomenon. The purpose of a CI system
> > is to be 99% green. In the 1% of cases where it fails, it should cause
> > development to stop so developers can research why, and fix a
> > regression before it goes into production or before new feature
> > development continues.
>
> Maybe you should investigate
> https://salsa.debian.org/salsa-ci-team/pipeline/-/commits/master?ref_type=heads
> which seems to be have a lot more than 1% failed pipelines.... :)
The git HEAD is and was green when you wrote this. As the this git
repo has no releases, only the HEAD passing all CI matters.
If you or others have CI enabled on projects where you can't fix the
repo so CI passes and git HEAD gets green, please scale down testing
scope until it passes, or disable the CI.
For the latest 1000 uploads to Debian, of those that have Salsa CI
enabled 32% were failing in the commit that was uploaded. That is not
good, and defeats the purpose of using a CI in the first place.
udd=> WITH latest AS (
SELECT source, ci_status
FROM vcswatch
ORDER BY last_scan DESC NULLS LAST
LIMIT 1000
)
SELECT
COUNT(*) AS total_packages,
COUNT(DISTINCT source) AS unique_packages,
COUNT(*) FILTER (WHERE ci_status = 'success') AS passing_ci,
COUNT(*) FILTER (WHERE ci_status <> 'success' AND ci_status IS NOT
NULL) AS failing_ci,
COUNT(*) FILTER (WHERE ci_status IS NULL) AS no_ci_status
FROM latest;
total_packages | unique_packages | passing_ci | failing_ci | no_ci_status
----------------+-----------------+------------+------------+--------------
1000 | 1000 | 375 | 177 | 448
