pulibrary / tigerdata-app / 6c0f9fdc-6aa1-470b-89ec-75afbb42fc46

94%

push

circleci

pulibraryrdss

This aims to propose optimizations to reduce execution times and improve
the efficiency of the CircleCI pipeline, transforming setup routines,
container readiness, cleanup, and test parallelism:

- Replace static sleep 10 timeouts with retry-based polling a command
sequence for PostgreSQL:
- Add wait_for_postgres to verify Postgres readiness using pg_isready
(up to 60s).
- Accelerate container cleanup in cleanup_docker by replacing slow,
global system/volume prunes with targeted stopping and deletion of
specific database/Mediaflux containers.
- Scale and parallelize RSpec test execution:
- Run the unit test job with parallelism of 4, filtering out system
specs and dynamically outputting segmented reports
(unit-${CIRCLE_NODE_INDEX}.xml).
- Run the test_browser job with parallelism of 2, targeting system and
view specs and dynamically outputting reports
(browser-${CIRCLE_NODE_INDEX}.xml).
- Leverage circleci tests split --split-by-timings to balance node
execution workloads.
- Avoid slow, redundant workspace persisting steps
(persist_to_workspace) across jobs.
- Clean up Node/Yarn tests (yarn_test job) by removing unnecessary
Chrome and Chromedriver installation steps, as Vitest runs entirely
within a lightweight headless jsdom environment.
- Skip redundant Vitest runs on the main mflux_ci pipeline since it is
already verified during the build phase's yarn_test job.
- Remove the 'validate' pre-commit hook from package.json.

0 of 314 branches covered (0.0%)

3310 of 3538 relevant lines covered (93.56%)

621.26 hits per line