Knapsack VS rspec_junit_formatter

This seemed extremely suspicious because we use Knapsack, which should ensure near-equal finishing times for all runners. Luckily, Knapsack stores run data and I was able to identify common offenders and identify the common thread - the spiking specs were writing to the database and then being cleaned up by DatabaseCleaner. The project had a complex DB setup, so I reached for the lowest-hanging fruit - I tried :deletion instead of :truncation cleanup strategy and it worked.

I've created a knapsack ruby gem for CI parallelisation that has over 122 million downloads. Primarily due to the fact, Gitlab is using it.

I spin off https://knapsackpro.com from the knapsack gem and we are helping our customers run fast CI builds.

name: Main on: [push] jobs: test: runs-on: ubuntu-latest # If you need DB like PostgreSQL, Redis then define service below. # https://github.com/actions/example-services/tree/master/.github/workflows services: postgres: image: postgres:10.8 env: POSTGRES_USER: postgres POSTGRES_PASSWORD: "" POSTGRES_DB: postgres ports: - 5432:5432 # needed because the postgres container does not provide a healthcheck # tmpfs makes DB faster by using RAM options: >- --mount type=tmpfs,destination=/var/lib/postgresql/data --health-cmd pg_isready --health-interval 10s --health-timeout 5s --health-retries 5 redis: image: redis ports: - 6379:6379 options: --entrypoint redis-server # https://help.github.com/en/articles/workflow-syntax-for-github-actions#jobsjob_idstrategymatrix strategy: fail-fast: false matrix: # [n] - where the n is a number of parallel jobs you want to run your tests on. # Use a higher number if you have slow tests to split them between more parallel jobs. # Remember to update the value of the `ci_node_index` below to (0..n-1). ci_node_total: [8] # Indexes for parallel jobs (starting from zero). # E.g. use [0, 1] for 2 parallel jobs, [0, 1, 2] for 3 parallel jobs, etc. ci_node_index: [0, 1, 2, 3, 4, 5, 6, 7] env: RAILS_ENV: test GEMFILE_RUBY_VERSION: 2.7.2 PGHOST: localhost PGUSER: postgres # Rails verifies the time zone in DB is the same as the time zone of the Rails app TZ: "Europe/Warsaw" steps: - uses: actions/checkout@v2 - name: Set up Ruby uses: ruby/setup-ruby@v1 with: # Not needed with a .ruby-version file ruby-version: 2.7 # runs 'bundle install' and caches installed gems automatically bundler-cache: true - name: Create DB run: | bin/rails db:prepare - name: Run tests env: KNAPSACK_PRO_TEST_SUITE_TOKEN_RSPEC: ${{ secrets.KNAPSACK_PRO_TEST_SUITE_TOKEN_RSPEC }} KNAPSACK_PRO_CI_NODE_TOTAL: ${{ matrix.ci_node_total }} KNAPSACK_PRO_CI_NODE_INDEX: ${{ matrix.ci_node_index }} KNAPSACK_PRO_LOG_LEVEL: info # if you use Knapsack Pro Queue Mode you must set below env variable # to be able to retry CI build and run previously recorded tests # https://github.com/KnapsackPro/knapsack_pro-ruby#knapsack_pro_fixed_queue_split-remember-queue-split-on-retry-ci-node KNAPSACK_PRO_FIXED_QUEUE_SPLIT: true # RSpec split test files by test examples feature - it's optional # https://knapsackpro.com/faq/question/how-to-split-slow-rspec-test-files-by-test-examples-by-individual-it KNAPSACK_PRO_RSPEC_SPLIT_BY_TEST_EXAMPLES: true run: | bundle exec rake knapsack_pro:queue:rspec

# Ruby CircleCI 2.0 configuration file # # Check https://circleci.com/docs/2.0/language-ruby/ for more details # version: 2 jobs: build: parallelism: 10 # https://circleci.com/docs/2.0/configuration-reference/#resource_class resource_class: small docker: # specify the version you desire here - image: circleci/ruby:2.7.1-node-browsers environment: PGHOST: 127.0.0.1 PGUSER: my_db_user RAILS_ENV: test # Split slow RSpec test files by test examples # https://knapsackpro.com/faq/question/how-to-split-slow-rspec-test-files-by-test-examples-by-individual-it KNAPSACK_PRO_RSPEC_SPLIT_BY_TEST_EXAMPLES: true # Specify service dependencies here if necessary # CircleCI maintains a library of pre-built images # documented at https://circleci.com/docs/2.0/circleci-images/ - image: circleci/postgres:10.6-alpine-ram environment: POSTGRES_DB: my_db_name POSTGRES_PASSWORD: "" POSTGRES_USER: my_db_user # Rails verifies Time Zone in DB is the same as time zone of the Rails app TZ: "Europe/Warsaw" - image: redis:6.0.7 working_directory: ~/repo environment: TZ: "Europe/Warsaw" steps: - checkout # Download and cache dependencies - restore_cache: keys: - v2-dependencies-bundler-{{ checksum "Gemfile.lock" }}-{{ checksum ".ruby-version" }} # fallback to using the latest cache if no exact match is found - v2-dependencies-bundler- - run: name: install ruby dependencies command: | bundle install --jobs=4 --retry=3 --path vendor/bundle - save_cache: paths: - ./vendor/bundle key: v2-dependencies-bundler-{{ checksum "Gemfile.lock" }}-{{ checksum ".ruby-version" }} # Database setup - run: bin/rails db:prepare - run: name: run tests command: | export CIRCLE_TEST_REPORTS=/tmp/test-results mkdir $CIRCLE_TEST_REPORTS bundle exec rake "knapsack_pro:queue:rspec[--format documentation --format RspecJunitFormatter --out tmp/rspec.xml]" # collect reports - store_test_results: path: /tmp/test-results - store_artifacts: path: /tmp/test-results destination: test-results

One, the interesting out-of-the-box solution is Knapsack Pro. It helps run your tests in a parallel efficient way.

rspec_junit_formatter - it’s a ruby gem that generates an XML report for executed tests with information about test failures. This report can be automatically read by CircleCI to present it in CircleCI web UI. No more browsing through long RSpec output - just look at highlighted failing specs in the TESTS tab :)