ci-backend.yml

# This workflow runs all of our backend django tests.
#
# If these tests get too slow, look at increasing concurrency and re-timing the tests by manually dispatching
# .github/workflows/ci-backend-update-test-timing.yml action
name: Backend CI

on:
    push:
        branches:
            - master
    workflow_dispatch:
        inputs:
            clickhouseServerVersion:
                description: ClickHouse server version. Leave blank for default
                type: string
    pull_request:

concurrency:
    group: ${{ github.workflow }}-${{ github.head_ref || github.run_id }}
    # This is so that the workflow run isn't canceled when a snapshot update is pushed within it by posthog-bot
    # We do however cancel from container-images-ci.yml if a commit is pushed by someone OTHER than posthog-bot
    cancel-in-progress: false

env:
    SECRET_KEY: '6b01eee4f945ca25045b5aab440b953461faf08693a9abbf1166dc7c6b9772da' # unsafe - for testing only
    DATABASE_URL: 'postgres://posthog:posthog@localhost:5432/posthog'
    REDIS_URL: 'redis://localhost'
    CLICKHOUSE_HOST: 'localhost'
    CLICKHOUSE_SECURE: 'False'
    CLICKHOUSE_VERIFY: 'False'
    TEST: 1
    CLICKHOUSE_SERVER_IMAGE_VERSION: ${{ github.event.inputs.clickhouseServerVersion || '' }}
    OBJECT_STORAGE_ENABLED: 'True'
    OBJECT_STORAGE_ENDPOINT: 'http://localhost:19000'
    OBJECT_STORAGE_ACCESS_KEY_ID: 'object_storage_root_user'
    OBJECT_STORAGE_SECRET_ACCESS_KEY: 'object_storage_root_password'
    # tests would intermittently fail in GH actions
    # with exit code 134 _after passing_ all tests
    # this appears to fix it
    # absolute wild tbh https://stackoverflow.com/a/75503402
    DISPLAY: ':99.0'
jobs:
    # Job to decide if we should run backend ci
    # See https://github.com/dorny/paths-filter#conditional-execution for more details
    changes:
        runs-on: ubuntu-latest
        timeout-minutes: 5
        name: Determine need to run backend checks
        # Set job outputs to values from filter step
        outputs:
            backend: ${{ steps.filter.outputs.backend }}
        steps:
            # For pull requests it's not necessary to checkout the code, but we
            # also want this to run on master so we need to checkout
            - uses: actions/checkout@v3

            - uses: dorny/paths-filter@v2
              id: filter
              with:
                  filters: |
                      backend:
                        # Avoid running backend tests for irrelevant changes
                        # NOTE: we are at risk of missing a dependency here. We could make
                        # the dependencies more clear if we separated the backend/frontend
                        # code completely
                        # really we should ignore ee/frontend/** but dorny doesn't support that
                        # - '!ee/frontend/**'
                        # including the negated rule appears to work
                        # but makes it always match because the checked file always isn't `ee/frontend/**` 🙈
                        - 'ee/**/*'
                        - 'hogvm/**/*'
                        - 'posthog/**/*'
                        - 'bin/*.py'
                        - requirements.txt
                        - requirements-dev.txt
                        - mypy.ini
                        - pytest.ini
                        - frontend/src/queries/schema.json # Used for generating schema.py
                        - plugin-transpiler/src # Used for transpiling plugins
                        # Make sure we run if someone is explicitly change the workflow
                        - .github/workflows/ci-backend.yml
                        - .github/actions/run-backend-tests/action.yml
                        # We use docker compose for tests, make sure we rerun on
                        # changes to docker-compose.dev.yml e.g. dependency
                        # version changes
                        - docker-compose.dev.yml
                        - frontend/public/email/*
                        # These scripts are used in the CI
                        - bin/check_temporal_up
                        - bin/check_kafka_clickhouse_up

    backend-code-quality:
        needs: changes
        timeout-minutes: 30

        name: Python code quality checks
        runs-on: ubuntu-latest

        steps:
            # If this run wasn't initiated by the bot (meaning: snapshot update) and we've determined
            # there are backend changes, cancel previous runs
            - uses: n1hility/cancel-previous-runs@v3
              if: github.actor != 'posthog-bot' && needs.changes.outputs.backend == 'true'
              with:
                  token: ${{ secrets.GITHUB_TOKEN }}

            - uses: actions/checkout@v3
              with:
                  fetch-depth: 1

            - name: Set up Python
              uses: actions/setup-python@v5
              with:
                  python-version: 3.11.9
                  cache: 'pip'
                  cache-dependency-path: '**/requirements*.txt'
                  token: ${{ secrets.POSTHOG_BOT_GITHUB_TOKEN }}

            # uv is a fast pip alternative: https://github.com/astral-sh/uv/
            - run: pip install uv

            - name: Install SAML (python3-saml) dependencies
              run: |
                  sudo apt-get update
                  sudo apt-get install libxml2-dev libxmlsec1 libxmlsec1-dev libxmlsec1-openssl

            - name: Install Python dependencies
              run: |
                  uv pip install --system -r requirements.txt -r requirements-dev.txt

            - name: Check for syntax errors, import sort, and code style violations
              run: |
                  ruff check .

            - name: Check formatting
              run: |
                  ruff format --check --diff .

            - name: Add Problem Matcher
              run: echo "::add-matcher::.github/mypy-problem-matcher.json"

            - name: Check static typing
              run: |
                  mypy -p posthog | mypy-baseline filter

            - name: Check if "schema.py" is up to date
              run: |
                  npm run schema:build:python && git diff --exit-code

    check-migrations:
        needs: changes
        if: needs.changes.outputs.backend == 'true'
        timeout-minutes: 10

        name: Validate Django and CH migrations
        runs-on: ubuntu-latest

        steps:
            - uses: actions/checkout@v3

            - name: Stop/Start stack with Docker Compose
              run: |
                  docker compose -f docker-compose.dev.yml down
                  docker compose -f docker-compose.dev.yml up -d

            - name: Set up Python
              uses: actions/setup-python@v5
              with:
                  python-version: 3.11.9
                  cache: 'pip'
                  cache-dependency-path: '**/requirements*.txt'
                  token: ${{ secrets.POSTHOG_BOT_GITHUB_TOKEN }}

            # uv is a fast pip alternative: https://github.com/astral-sh/uv/
            - run: pip install uv

            - name: Install SAML (python3-saml) dependencies
              run: |
                  sudo apt-get update
                  sudo apt-get install libxml2-dev libxmlsec1-dev libxmlsec1-openssl

            # First running migrations from master, to simulate the real-world scenario
            - name: Checkout master
              uses: actions/checkout@v3
              with:
                  ref: master

            - name: Install python dependencies for master
              run: |
                  uv pip install --system -r requirements.txt -r requirements-dev.txt

            - name: Run migrations up to master
              run: |
                  python manage.py migrate

            # Now we can consider this PR's migrations
            - name: Checkout this PR
              uses: actions/checkout@v3

            - name: Install python dependencies for this PR
              run: |
                  uv pip install --system -r requirements.txt -r requirements-dev.txt

            - name: Run migrations for this PR
              run: |
                  python manage.py migrate

            - name: Check migrations
              run: |
                  python manage.py makemigrations --check --dry-run
                  git fetch origin master
                  # `git diff --name-only` returns a list of files that were changed - added OR deleted OR modified
                  # With `--name-status` we get the same, but including a column for status, respectively: A, D, M
                  # In this check we exclusively care about files that were
                  # added (A) in posthog/migrations/. We also want to ignore
                  # initial migrations (0001_*) as these are guaranteed to be
                  # run on initial setup where there is no data.
                  git diff --name-status origin/master..HEAD | grep "A\sposthog/migrations/" | awk '{print $2}' | grep -v migrations/0001_ | python manage.py test_migrations_are_safe

            - name: Check CH migrations
              run: |
                  # Same as above, except now for CH looking at files that were added in posthog/clickhouse/migrations/
                  git diff --name-status origin/master..HEAD | grep "A\sposthog/clickhouse/migrations/" | awk '{print $2}' |  python manage.py test_ch_migrations_are_safe

    django:
        needs: changes
        # increase for tmate testing
        timeout-minutes: 30

        name: Django tests – ${{ matrix.segment }} (persons-on-events ${{ matrix.person-on-events && 'on' || 'off' }}), Py ${{ matrix.python-version }}, ${{ matrix.clickhouse-server-image }} (${{matrix.group}}/${{ matrix.concurrency }})
        runs-on: ubuntu-latest

        strategy:
            fail-fast: false
            matrix:
                python-version: ['3.11.9']
                clickhouse-server-image: ['clickhouse/clickhouse-server:23.12.6.19']
                segment: ['Core']
                person-on-events: [false, true]
                # :NOTE: Keep concurrency and groups in sync
                concurrency: [10]
                group: [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]
                include:
                    - segment: 'Temporal'
                      person-on-events: false
                      clickhouse-server-image: 'clickhouse/clickhouse-server:23.12.6.19'
                      python-version: '3.11.9'
                      concurrency: 3
                      group: 1
                    - segment: 'Temporal'
                      person-on-events: false
                      clickhouse-server-image: 'clickhouse/clickhouse-server:23.12.6.19'
                      python-version: '3.11.9'
                      concurrency: 3
                      group: 2
                    - segment: 'Temporal'
                      person-on-events: false
                      clickhouse-server-image: 'clickhouse/clickhouse-server:23.12.6.19'
                      python-version: '3.11.9'
                      concurrency: 3
                      group: 3

        steps:
            # The first step is the only one that should run if `needs.changes.outputs.backend == 'false'`.
            # All the other ones should rely on `needs.changes.outputs.backend` directly or indirectly, so that they're
            # effectively skipped if backend code is unchanged. See https://github.com/PostHog/posthog/pull/15174.
            - uses: actions/checkout@v3
              with:
                  fetch-depth: 1
                  repository: ${{ github.event.pull_request.head.repo.full_name }}
                  ref: ${{ github.event.pull_request.head.ref }}
                  # Use PostHog Bot token when not on forks to enable proper snapshot updating
                  token: ${{ github.event.pull_request.head.repo.full_name == github.repository && secrets.POSTHOG_BOT_GITHUB_TOKEN || github.token }}

            - uses: ./.github/actions/run-backend-tests
              if: needs.changes.outputs.backend == 'true'
              with:
                  segment: ${{ matrix.segment }}
                  person-on-events: ${{ matrix.person-on-events }}
                  python-version: ${{ matrix.python-version }}
                  clickhouse-server-image: ${{ matrix.clickhouse-server-image }}
                  concurrency: ${{ matrix.concurrency }}
                  group: ${{ matrix.group }}
                  token: ${{ secrets.POSTHOG_BOT_GITHUB_TOKEN }}

            - uses: EndBug/add-and-commit@v9
              # Skip on forks
              # Also skip for persons-on-events runs, as we want to ignore snapshots diverging there
              if: ${{ github.event.pull_request.head.repo.full_name == 'PostHog/posthog' && needs.changes.outputs.backend == 'true' && !matrix.person-on-events }}
              with:
                  add: '["ee", "./**/*.ambr", "posthog/queries/", "posthog/migrations", "posthog/tasks", "posthog/hogql/"]'
                  message: 'Update query snapshots'
                  pull: --rebase --autostash # Make sure we're up-to-date with other segments' updates
                  default_author: github_actions
                  github_token: ${{ secrets.POSTHOG_BOT_GITHUB_TOKEN }}

            - name: Check if any snapshot changes were left uncomitted
              id: changed-files
              if: ${{ github.event.pull_request.head.repo.full_name == 'PostHog/posthog' && needs.changes.outputs.backend == 'true' && !matrix.person-on-events }}
              run: |
                  if [[ -z $(git status -s | grep -v ".test_durations" | tr -d "\n") ]]
                  then
                    echo 'files_found=false' >> $GITHUB_OUTPUT
                  else
                    echo 'diff=$(git status --porcelain)' >> $GITHUB_OUTPUT
                    echo 'files_found=true' >> $GITHUB_OUTPUT
                  fi

            - name: Fail CI if some snapshots have been updated but not committed
              if: steps.changed-files.outputs.files_found == 'true' && steps.add-and-commit.outcome == 'success'
              run: |
                  echo "${{ steps.changed-files.outputs.diff }}"
                  exit 1

            - name: Archive email renders
              uses: actions/upload-artifact@v3
              if: needs.changes.outputs.backend == 'true' && matrix.segment == 'Core' && matrix.person-on-events == false
              with:
                  name: email_renders
                  path: posthog/tasks/test/__emails__
                  retention-days: 5

    async-migrations:
        name: Async migrations tests -  ${{ matrix.clickhouse-server-image }}
        needs: changes
        strategy:
            fail-fast: false
            matrix:
                clickhouse-server-image: ['clickhouse/clickhouse-server:23.12.6.19']
        if: needs.changes.outputs.backend == 'true'
        runs-on: ubuntu-latest
        steps:
            - name: 'Checkout repo'
              uses: actions/checkout@v3
              with:
                  fetch-depth: 1

            - name: Start stack with Docker Compose
              run: |
                  export CLICKHOUSE_SERVER_IMAGE_VERSION=${{ matrix.clickhouse-server-image }}
                  docker compose -f docker-compose.dev.yml down
                  docker compose -f docker-compose.dev.yml up -d

            - name: Set up Python
              uses: actions/setup-python@v5
              with:
                  python-version: 3.11.9
                  cache: 'pip'
                  cache-dependency-path: '**/requirements*.txt'
                  token: ${{ secrets.POSTHOG_BOT_GITHUB_TOKEN }}

            # uv is a fast pip alternative: https://github.com/astral-sh/uv/
            - run: pip install uv

            - name: Install SAML (python3-saml) dependencies
              run: |
                  sudo apt-get update
                  sudo apt-get install libxml2-dev libxmlsec1-dev libxmlsec1-openssl

            - name: Install python dependencies
              shell: bash
              run: |
                  uv pip install --system -r requirements.txt -r requirements-dev.txt

            - name: Add kafka host to /etc/hosts for kafka connectivity
              run: sudo echo "127.0.0.1 kafka" | sudo tee -a /etc/hosts

            - name: Set up needed files
              run: |
                  mkdir -p frontend/dist
                  touch frontend/dist/index.html
                  touch frontend/dist/layout.html
                  touch frontend/dist/exporter.html

            - name: Wait for Clickhouse & Kafka
              run: bin/check_kafka_clickhouse_up

            - name: Run async migrations tests
              run: |
                  pytest -m "async_migrations"

    calculate-running-time:
        name: Calculate running time
        needs: [django, async-migrations]
        runs-on: ubuntu-latest
        if: # Run on pull requests to PostHog/posthog + on PostHog/posthog outside of PRs - but never on forks
            needs.changes.outputs.backend == 'true' &&
            (
            github.event_name == 'pull_request' && github.event.pull_request.head.repo.full_name || github.repository
            ) == 'PostHog/posthog'
        steps:
            - name: Calculate running time
              run: |
                  echo "${{ secrets.GITHUB_TOKEN }}" | gh auth login --with-token
                  run_id=${GITHUB_RUN_ID}
                  repo=${GITHUB_REPOSITORY}
                  run_info=$(gh api repos/${repo}/actions/runs/${run_id})
                  echo run_info: ${run_info}
                  # name is the name of the workflow file
                  # run_started_at is the start time of the workflow
                  # we want to get the number of seconds between the start time and now
                  name=$(echo ${run_info} | jq -r '.name')
                  run_url=$(echo ${run_info} | jq -r '.url')
                  run_started_at=$(echo ${run_info} | jq -r '.run_started_at')
                  run_attempt=$(echo ${run_info} | jq -r '.run_attempt')
                  start_seconds=$(date -d "${run_started_at}" +%s)
                  now_seconds=$(date +%s)
                  duration=$((now_seconds-start_seconds))
                  echo running_time_duration_seconds=${duration} >> $GITHUB_ENV
                  echo running_time_run_url=${run_url} >> $GITHUB_ENV
                  echo running_time_run_attempt=${run_attempt} >> $GITHUB_ENV
                  echo running_time_run_id=${run_id} >> $GITHUB_ENV
                  echo running_time_run_started_at=${run_started_at} >> $GITHUB_ENV
            - name: Capture running time to PostHog
              uses: PostHog/posthog-github-action@v0.1
              with:
                  posthog-token: ${{secrets.POSTHOG_API_TOKEN}}
                  event: 'posthog-ci-running-time'
                  properties: '{"duration_seconds": ${{ env.running_time_duration_seconds }}, "run_url": "${{ env.running_time_run_url }}", "run_attempt": "${{ env.running_time_run_attempt }}", "run_id": "${{ env.running_time_run_id }}", "run_started_at": "${{ env.running_time_run_started_at }}"}'