Docker Compose¶

docker compose logs interleaves output from every service. The result is noisy, repetitive, and full of timestamps and container IDs - exactly the shape LogStrip is built for.

Snapshot all services¶

docker compose logs --no-color > raw_logs.txt
npx -y logstrip raw_logs.txt -o clean.log --stats

Pass --no-color so ANSI escapes do not leak into the compressed output.

One step with stdin¶

docker compose logs --no-color 2>&1 | npx -y logstrip > clean.log

Follow mode with a time budget¶

When tailing live logs, cap the run so the CLI exits cleanly:

docker compose logs --no-color --follow 2>&1 \
  | npx -y logstrip --timeout 60 -o clean.log --stats

--timeout 60 flushes the compressed output after 60 seconds and sets timedOut: true in --json output. See CLI reference - sampling and timeouts.

Per-service compression¶

When one service dominates the noise, compress it on its own:

docker compose logs --no-color api > raw_api.log
docker compose logs --no-color worker > raw_worker.log

npx -y logstrip raw_api.log    -o clean_api.log    --stats
npx -y logstrip raw_worker.log -o clean_worker.log --stats

CI: capture compose logs on failure¶

- name: Bring up stack
  run: docker compose up -d --wait

- name: Run integration tests
  run: pytest tests/integration -ra
  continue-on-error: true
  id: tests

- name: Capture compose logs on failure
  if: steps.tests.outcome == 'failure'
  run: |
    docker compose logs --no-color > raw_logs.txt
    npx -y logstrip raw_logs.txt -o clean.log --stats

- name: Upload compact logs
  if: steps.tests.outcome == 'failure'
  uses: actions/upload-artifact@v4
  with:
    name: compose-logs
    path: clean.log

Tighter compression for chatty stacks¶

Stacks with health-check spam (databases, brokers, gateways) benefit from high or aggressive:

docker compose logs --no-color \
  | npx -y logstrip -a aggressive --exclude 'healthcheck|ping' > clean.log

See CLI reference - aggressiveness and include/exclude patterns.