discourse/.github/workflows/tests.yml
Alan Guo Xiang Tan 39da9106ba
DEV: Introduce automatic reruns to RSpec tests on Github actions (#24811)
What motivated this change?

Our builds on Github actions have been extremely flaky mostly due to system tests. This has led to a drop in confidence
in our test suite where our developers tend to assume that a failed job is due to a flaky system test. As a result, we
have had occurrences where changes that resulted in legitimate test failures are merged into the `main` branch because developers
assumed it was a flaky test.

What does this change do?

This change seeks to reduce the flakiness of our builds on Github Actions by automatically re-running RSpec tests once when
they fail. If a failed test passes subsequently in the re-run, we mark the test as flaky by logging it into a file on disk
which is then uploaded as an artifact of the Github workflow run. We understand that automatically re-runs will lead to 
lower accuracy of our tests but we accept this as an acceptable trade-off since a fragile build has a much greater impact
on our developers' time. Internally, the Discourse development team will be running a service to fetch the flaky tests 
which have been logged for internal monitoring.

How is the change implemented?

1. A `--retry-and-log-flaky-tests` CLI flag is added to the `bin/turbo_rspec` CLI which will then initialize `TurboTests::Runner` 
with the `retry_and_log_flaky_tests` kwarg set to `true`. 

2. When the `retry_and_log_flaky_tests` kwarg is set to `true` for `TurboTests::Runner`, we will register an additional 
formatter `Flaky::FailuresLoggerFormatter` to the `TurboTests::Reporter` in the `TurboTests::Runner#run` method. 
The `Flaky::FailuresLoggerFormatter` has a simple job of logging all failed examples to a file on disk when running all the 
tests. The details of the failed example which are logged can be found in `TurboTests::Flaky::FailedExample.to_h`.

3. Once all the tests have been run once, we check the result for any failed examples and if there are, we read the file on
disk to fetch the `location_rerun_location` of the failed examples which is then used to run the tests in a new RSpec process.
In the rerun, we configure a `TurboTests::Flaky::FlakyDetectorFormatter` with RSpec which removes all failed examples from the log file on disk since those examples are not flaky tests. Note that if there are too many failed examples on the first run, we will deem the failures to likely not be due to flaky tests and not re-run the test failures. As of writing, the threshold of failed examples is set to 10. If there are more than 10 failed examples, we will not re-run the failures.
2023-12-13 07:18:27 +08:00

415 lines
15 KiB
YAML

name: Tests
on:
pull_request:
paths-ignore:
- "migrations/**"
push:
branches:
- main
- beta
- stable
paths-ignore:
- "migrations/**"
concurrency:
group: tests-${{ format('{0}-{1}', github.head_ref || github.run_number, github.job) }}
cancel-in-progress: true
permissions:
contents: read
jobs:
build:
if: github.event_name == 'pull_request' || github.repository != 'discourse/discourse-private-mirror'
name: ${{ matrix.target }} ${{ matrix.build_type }}${{ matrix.updated_ember && ' (Ember 5)' || '' }}
runs-on: ${{ (matrix.build_type == 'annotations') && 'ubuntu-latest' || 'ubuntu-20.04-8core' }}
container: discourse/discourse_test:slim${{ (matrix.build_type == 'frontend' || matrix.build_type == 'system') && '-browsers' || '' }}${{ (matrix.ruby == '3.1') && '-ruby-3.1.0' || '' }}
timeout-minutes: 20
env:
DISCOURSE_HOSTNAME: www.example.com
RAILS_ENV: test
PGUSER: discourse
PGPASSWORD: discourse
USES_PARALLEL_DATABASES: ${{ matrix.build_type == 'backend' || matrix.build_type == 'system' }}
CAPYBARA_DEFAULT_MAX_WAIT_TIME: 10
MINIO_RUNNER_LOG_LEVEL: DEBUG
EMBER_VERSION: ${{ (matrix.updated_ember && '5') || '3' }}
strategy:
fail-fast: false
matrix:
build_type: [backend, frontend, system, annotations]
target: [core, plugins, themes]
ruby: ["3.2"]
updated_ember: [false]
exclude:
- build_type: annotations
target: plugins
- build_type: annotations
target: themes
- build_type: backend
target: themes
- build_type: frontend
target: core # Handled by core_frontend_tests job (below)
include:
- build_type: system
target: chat
- build_type: system
target: chat
updated_ember: true
- build_type: system
target: core
updated_ember: true
- build_type: frontend
target: plugins
updated_ember: true
- build_type: frontend
target: themes
updated_ember: true
steps:
- name: Set working directory owner
run: chown root:root .
- uses: actions/checkout@v4
with:
fetch-depth: 1
- name: Setup Git
run: |
git config --global user.email "ci@ci.invalid"
git config --global user.name "Discourse CI"
- name: Start redis
run: |
redis-server /etc/redis/redis.conf &
- name: Start Postgres
run: |
chown -R postgres /var/run/postgresql
sudo -E -u postgres script/start_test_db.rb
sudo -u postgres psql -c "CREATE ROLE $PGUSER LOGIN SUPERUSER PASSWORD '$PGPASSWORD';"
- name: Bundler cache
uses: actions/cache@v3
with:
path: vendor/bundle
key: ${{ runner.os }}-${{ matrix.ruby }}-gem-${{ hashFiles('**/Gemfile.lock') }}-cachev2
- name: Setup gems
run: |
gem install bundler --conservative -v $(awk '/BUNDLED WITH/ { getline; gsub(/ /,""); print $0 }' Gemfile.lock)
bundle config --local path vendor/bundle
bundle config --local deployment true
bundle config --local without development
bundle install --jobs 4
bundle clean
- name: Get yarn cache directory
id: yarn-cache-dir
run: echo "dir=$(yarn cache dir)" >> $GITHUB_OUTPUT
- name: Yarn cache
uses: actions/cache@v3
id: yarn-cache
with:
path: ${{ steps.yarn-cache-dir.outputs.dir }}
key: ${{ runner.os }}-yarn-${{ hashFiles('**/yarn.lock') }}-cachev2
- name: Upgrade Ember
if: matrix.updated_ember == true
run: script/switch_ember_version 5
- name: Yarn install
run: yarn install --frozen-lockfile
- name: Checkout official plugins
if: matrix.target == 'plugins'
run: bin/rake plugin:install_all_official
- name: Pull compatible versions of plugins
if: matrix.target == 'plugins'
run: bin/rake plugin:pull_compatible_all
- name: Plugin gems cache
uses: actions/cache@v3
with:
path: plugins/*/gems
key: ${{ runner.os }}-plugin-gems-${{matrix.ruby}}-${{ hashFiles('plugins/*/plugin.rb') }}
- name: Checkout official themes
if: matrix.target == 'themes' && matrix.build_type == 'system'
run: bin/rake themes:clone_all_official
- name: Add hosts to /etc/hosts, otherwise Chrome cannot reach minio
if: matrix.build_type == 'system' && matrix.target == 'core'
run: |
echo "127.0.0.1 minio.local" | sudo tee -a /etc/hosts
echo "127.0.0.1 discoursetest.minio.local" | sudo tee -a /etc/hosts
- name: Fetch app state cache
uses: actions/cache@v3
id: app-cache
with:
path: tmp/app-cache
key: >-
${{ runner.os }}-
${{ hashFiles('.github/workflows/tests.yml') }}-
${{ hashFiles('db/**/*', 'plugins/**/db/**/*') }}-
${{ env.USES_PARALLEL_DATABASES }}
- name: Restore database from cache
if: steps.app-cache.outputs.cache-hit == 'true'
run: script/silence_successful_output psql --quiet -o /dev/null -f tmp/app-cache/cache.sql postgres
- name: Restore uploads from cache
if: steps.app-cache.outputs.cache-hit == 'true'
run: rm -rf public/uploads && cp -r tmp/app-cache/uploads public/uploads
- name: Create and migrate database
if: steps.app-cache.outputs.cache-hit != 'true'
run: |
bin/rake db:create
script/silence_successful_output bin/rake db:migrate
- name: Create and migrate parallel databases
if: >-
env.USES_PARALLEL_DATABASES == 'true' &&
steps.app-cache.outputs.cache-hit != 'true'
run: |
bin/rake parallel:create
script/silence_successful_output bin/rake parallel:migrate
- name: Dump database for cache
if: steps.app-cache.outputs.cache-hit != 'true'
run: mkdir -p tmp/app-cache && pg_dumpall > tmp/app-cache/cache.sql
- name: Dump uploads for cache
if: steps.app-cache.outputs.cache-hit != 'true'
run: rm -rf tmp/app-cache/uploads && cp -r public/uploads tmp/app-cache/uploads
- name: Fetch turbo_rspec_runtime.log cache
uses: actions/cache@v3
id: test-runtime-cache
if: matrix.build_type == 'backend' || matrix.build_type == 'system'
with:
path: tmp/turbo_rspec_runtime.log
key: rspec-runtime-${{ matrix.build_type }}-${{ matrix.target }}-${{ github.run_id }}
restore-keys: rspec-runtime-${{ matrix.build_type }}-${{ matrix.target }}-
- name: Check Zeitwerk eager_load
if: matrix.build_type == 'backend'
env:
LOAD_PLUGINS: ${{ (matrix.target == 'plugins') && '1' || '0' }}
run: |
if ! bin/rails zeitwerk:check --trace; then
echo
echo "---------------------------------------------"
echo
echo "::error::'bin/rails zeitwerk:check' failed - the app will fail to boot with 'eager_load=true' (e.g. in production)."
echo "To reproduce locally, run 'bin/rails zeitwerk:check'."
echo "Alternatively, you can run your local server/tests with the 'DISCOURSE_ZEITWERK_EAGER_LOAD=1' environment variable."
echo
exit 1
fi
- name: Check Zeitwerk reloading
if: matrix.build_type == 'backend'
env:
LOAD_PLUGINS: ${{ (matrix.target == 'plugins') && '1' || '0' }}
run: |
if ! bin/rails runner 'Rails.application.reloader.reload!'; then
echo
echo "---------------------------------------------"
echo
echo "::error::Zeitwerk reload failed - the app will not be able to reload properly in development."
echo "To reproduce locally, run \`bin/rails runner 'Rails.application.reloader.reload!'\`."
echo
exit 1
fi
- name: Core RSpec
if: matrix.build_type == 'backend' && matrix.target == 'core'
run: bin/turbo_rspec --use-runtime-info --verbose --format documentation --retry-and-log-flaky-tests
- name: Plugin RSpec
if: matrix.build_type == 'backend' && matrix.target == 'plugins'
run: bin/rake plugin:turbo_spec['*','--verbose --format documentation --use-runtime-info --retry-and-log-flaky-tests']
- name: Plugin QUnit
if: matrix.build_type == 'frontend' && matrix.target == 'plugins'
run: QUNIT_WRITE_EXECUTION_FILE=1 QUNIT_PARALLEL=3 bin/rake plugin:qunit['*','1200000']
timeout-minutes: 30
- name: Theme QUnit
if: matrix.build_type == 'frontend' && matrix.target == 'themes'
run: DISCOURSE_DEV_DB=discourse_test QUNIT_PARALLEL=3 bin/rake themes:qunit_all_official
timeout-minutes: 15
- uses: actions/upload-artifact@v3
if: always() && matrix.build_type == 'frontend' && matrix.target == 'plugins'
with:
name: ember-exam-execution-plugins-frontend
path: ./app/assets/javascripts/discourse/test-execution-*.json
- name: Ember Build for System Tests
if: matrix.build_type == 'system'
run: bin/ember-cli --build
- name: Ensure latest minio binary installed for Core System Tests
if: matrix.build_type == 'system' && matrix.target == 'core'
run: bundle exec ruby script/install_minio_binaries.rb
- name: Core System Tests
if: matrix.build_type == 'system' && matrix.target == 'core'
run: RAILS_ENABLE_TEST_LOG=1 RAILS_TEST_LOG_LEVEL=error PARALLEL_TEST_PROCESSORS=4 bin/turbo_rspec --use-runtime-info --profile=50 --verbose --format documentation --retry-and-log-flaky-tests spec/system
- name: Plugin System Tests
if: matrix.build_type == 'system' && matrix.target == 'plugins'
run: |
GLOBIGNORE="plugins/chat/*";
LOAD_PLUGINS=1 RAILS_ENABLE_TEST_LOG=1 RAILS_TEST_LOG_LEVEL=error PARALLEL_TEST_PROCESSORS=4 bin/turbo_rspec --use-runtime-info --profile=50 --verbose --format documentation --retry-and-log-flaky-tests plugins/*/spec/system
shell: bash
timeout-minutes: 30
- name: Chat System Tests
if: matrix.build_type == 'system' && matrix.target == 'chat'
run: LOAD_PLUGINS=1 RAILS_ENABLE_TEST_LOG=1 RAILS_TEST_LOG_LEVEL=error PARALLEL_TEST_PROCESSORS=4 bin/turbo_rspec --use-runtime-info --profile=50 --verbose --format documentation --retry-and-log-flaky-tests plugins/chat/spec/system
timeout-minutes: 30
- name: Theme System Tests
if: matrix.build_type == 'system' && matrix.target == 'themes'
run: |
RAILS_ENABLE_TEST_LOG=1 RAILS_TEST_LOG_LEVEL=error PARALLEL_TEST_PROCESSORS=4 bin/turbo_rspec --profile=50 --verbose --format documentation --retry-and-log-flaky-tests tmp/themes/*/spec/system
shell: bash
timeout-minutes: 30
- name: Check for failed system test screenshots
id: check-failed-system-test-screenshots
if: always() && matrix.build_type == 'system'
run: |
if [ -d tmp/capybara ] && [ "$(ls -A tmp/capybara/)" ]; then
echo "exists=true" >> $GITHUB_OUTPUT
else
echo "exists=false" >> $GITHUB_OUTPUT
fi
shell: bash
- name: Upload failed system test screenshots
uses: actions/upload-artifact@v3
if: always() && steps.check-failed-system-test-screenshots.outputs.exists == 'true'
with:
name: failed-system-test-screenshots
path: tmp/capybara/*.png
- name: Check for flaky tests report
id: check-flaky-spec-report
if: always()
run: |
if [ -f tmp/turbo_rspec_flaky_tests.json ]; then
echo "exists=true" >> $GITHUB_OUTPUT
else
echo "exists=false" >> $GITHUB_OUTPUT
fi
shell: bash
- name: Create flaky tests report artifact
if: always() && steps.check-flaky-spec-report.outputs.exists == 'true'
run: cp tmp/turbo_rspec_flaky_tests.json tmp/turbo_rspec_flaky_tests-${{ matrix.build_type }}-${{ matrix.target }}.json
- name: Upload flaky tests report
uses: actions/upload-artifact@v3
if: always() && steps.check-flaky-spec-report.outputs.exists == 'true'
with:
name: flaky-test-reports
path: tmp/turbo_rspec_flaky_tests-${{ matrix.build_type }}-${{ matrix.target }}.json
- name: Check Annotations
if: matrix.build_type == 'annotations'
run: |
bin/rake annotate:ensure_all_indexes
bin/annotate --models --model-dir app/models
if [ ! -z "$(git status --porcelain app/models/)" ]; then
echo "Core annotations are not up to date. To resolve, run:"
echo " bin/rake annotate:clean"
echo
echo "Or manually apply the diff printed below:"
echo "---------------------------------------------"
git -c color.ui=always diff app/models/
exit 1
fi
timeout-minutes: 30
core_frontend_tests:
if: github.event_name == 'pull_request' || github.repository != 'discourse/discourse-private-mirror'
name: core frontend (${{ matrix.browser }})${{ matrix.updated_ember && ' (Ember 5)' || '' }}
runs-on: ubuntu-20.04-8core
container:
image: discourse/discourse_test:slim-browsers
options: --user discourse
timeout-minutes: 35
strategy:
fail-fast: false
matrix:
browser: ["Chrome", "Firefox ESR", "Firefox Evergreen"]
updated_ember: [false]
include:
- browser: Chrome
updated_ember: true
env:
TESTEM_BROWSER: ${{ (startsWith(matrix.browser, 'Firefox') && 'Firefox') || matrix.browser }}
TESTEM_FIREFOX_PATH: ${{ (matrix.browser == 'Firefox Evergreen') && '/opt/firefox-evergreen/firefox' }}
steps:
- uses: actions/checkout@v4
with:
fetch-depth: 1
- name: Setup Git
run: |
git config --global user.email "ci@ci.invalid"
git config --global user.name "Discourse CI"
- name: Get yarn cache directory
id: yarn-cache-dir
run: echo "dir=$(yarn cache dir)" >> $GITHUB_OUTPUT
- name: Yarn cache
uses: actions/cache@v3
id: yarn-cache
with:
path: ${{ steps.yarn-cache-dir.outputs.dir }}
key: ${{ runner.os }}-yarn-${{ hashFiles('**/yarn.lock') }}-cachev2
- name: Upgrade Ember
if: matrix.updated_ember == true
run: script/switch_ember_version 5
- name: Yarn install
working-directory: ./app/assets/javascripts/discourse
run: yarn install --frozen-lockfile
- name: Ember Build
working-directory: ./app/assets/javascripts/discourse
run: |
mkdir /tmp/emberbuild
yarn ember build --environment=test -o /tmp/emberbuild
- name: Core QUnit
working-directory: ./app/assets/javascripts/discourse
run: yarn ember exam --path /tmp/emberbuild --load-balance --parallel=5 --launch "${{ env.TESTEM_BROWSER }}" --write-execution-file --random
timeout-minutes: 15
- uses: actions/upload-artifact@v3
if: ${{ always() }}
with:
name: ember-exam-execution-${{ matrix.browser }}
path: ./app/assets/javascripts/discourse/test-execution-*.json