How to use this image

All examples in this guide use the public image. If you've mirrored the repository for your own use (for example, to your Docker Hub namespace), update your commands to reference the mirrored image instead of the public one.

For example:

• Public image: dhi.io/<repository>:<tag>
• Mirrored image: <your-namespace>/dhi-<repository>:<tag>

For the examples, you must first use docker login dhi.io to authenticate to the registry to pull the images.

Prerequisites

Codecov API is one service of a self-hosted Codecov deployment and needs the rest of the stack at runtime:

• PostgreSQL for the main database.
• TimescaleDB (PostgreSQL with the timescaledb extension) for the timeseries databases, when setup.timeseries or setup.ta_timeseries is enabled in the install configuration.
• Redis for caching and for the task queue shared with the Codecov worker.
• An S3-compatible object store (MinIO, S3 or GCS) for coverage reports.
• The Codecov worker, frontend and gateway services to complete a deployment.

The image bundles none of these. The Codecov self-hosted repository⁠ carries the reference compose file and install configuration this image follows, and the Codecov configuration reference⁠ documents every configuration key.

Start a Codecov API instance

Codecov reads its install configuration from /config/codecov.yml, or from the path in CODECOV_YML. Any key can also be set through the environment as SECTION__KEY (for example SERVICES__DATABASE_URL). A minimal config/codecov.yml for a local stack:

setup:
  codecov_url: http://localhost:8000
  enterprise_license: "<your-license-key>"
  http:
    cookie_secret: "<random-string>"
  timeseries:
    enabled: true
  ta_timeseries:
    enabled: true
services:
  redis_url: redis://redis:6379
  database_url: postgres://postgres:testpassword@postgres:5432/postgres
  timeseries_database_url: postgres://postgres:testpassword@timescale:5432/postgres
  ta_timeseries_database_url: postgres://postgres:testpassword@timescale:5432/postgres
django:
  secret_key: "<random-string>"

Django's SECRET_KEY defaults to a fixed value when django.secret_key is unset, so set it. Without a valid enterprise_license the API starts, but the licence resolves as invalid with zero seats and no user can be activated; upstream treats the key as required. Run the API next to its data stores with Compose:

services:
  api:
    image: dhi.io/codecov-api:<tag>
    ports:
      - "127.0.0.1:8000:8000"
    volumes:
      - ./config:/config
    depends_on:
      postgres:
        condition: service_healthy
      timescale:
        condition: service_healthy
      redis:
        condition: service_healthy

  postgres:
    image: docker.io/postgres:17
    environment:
      POSTGRES_USER: postgres
      POSTGRES_PASSWORD: testpassword
      POSTGRES_DB: postgres
    volumes:
      - postgres-data:/var/lib/postgresql/data
    healthcheck:
      test: ["CMD", "pg_isready", "-U", "postgres", "-d", "postgres"]
      interval: 5s
      retries: 20

  timescale:
    image: docker.io/timescale/timescaledb:latest-pg17
    environment:
      POSTGRES_USER: postgres
      POSTGRES_PASSWORD: testpassword
      POSTGRES_DB: postgres
    volumes:
      - timescale-data:/var/lib/postgresql/data
    healthcheck:
      test: ["CMD", "pg_isready", "-U", "postgres", "-d", "postgres"]
      interval: 5s
      retries: 20

  redis:
    image: docker.io/redis:7
    volumes:
      - redis-data:/data
    healthcheck:
      test: ["CMD", "redis-cli", "ping"]
      interval: 5s
      retries: 20

volumes:
  postgres-data:
  timescale-data:
  redis-data:

When RUN_ENV is ENTERPRISE (the default) or DEV, the entrypoint applies the Django migrations, the timeseries migrations and the PostgreSQL partition setup on every start unless CODECOV_SKIP_MIGRATIONS=true; other RUN_ENV values skip them. It then starts gunicorn on port 8000. GET http://localhost:8000/health/ answers HTTP 200 with a body ending in is live! once the API is serving. The API port is bound to the loopback interface above because the service is meant to sit behind the Codecov gateway; expose it directly only on a trusted network.

Environment variables

The entrypoint honours these variables; install-configuration keys are read from the configuration file or SECTION__KEY variables.

Common Codecov API use cases

Run a management command

Passing a command to the image runs it inside the application environment instead of starting gunicorn. Skip the entrypoint's own migration step when the command should run on its own:

$ docker run --rm --network <stack-network> -v ./config:/config \
    -e CODECOV_SKIP_MIGRATIONS=true \
    dhi.io/codecov-api:<tag> python manage.py showmigrations

The same form runs the migrations as a separate job before rolling out a new API version:

$ docker run --rm --network <stack-network> -v ./config:/config \
    -e CODECOV_SKIP_MIGRATIONS=true \
    dhi.io/codecov-api:<tag> bash -c './migrate.sh && ./migrate-timeseries.sh'

migrate.sh and migrate-timeseries.sh come from upstream and are executable in this image. They run the Django migrations and partition setup and the timeseries migrations, the same steps the entrypoint runs.

Validate a repository codecov.yml

The API validates a repository codecov.yml (not the install configuration) without authentication:

$ curl -X POST --data-binary @- http://localhost:8000/validate <<'EOF'
coverage:
  status:
    project:
      default:
        target: 80%
EOF

Health checks and metrics

/health/ returns HTTP 200 and a body ending in is live! when the API can reach its database, and /monitoring/metrics exposes Prometheus metrics. Runtime images carry no curl, so probe from the orchestrator:

readinessProbe:
  httpGet:
    path: /health/
    port: 8000
  periodSeconds: 5

FIPS scope

The FIPS variants load the OpenSSL FIPS provider for the system libcrypto. Python's ssl, hashlib and hmac modules, the system libpq behind psycopg2 and the cryptography package (built from source against the system OpenSSL rather than the wheel's bundled copy) all use it, so Django's password hashing and session signing, outbound HTTPS through requests, httpx and urllib3, TLS to PostgreSQL, JWT signing, the AES encryption of the enterprise licence and of stored VCS tokens, and X.509 handling run through the validated module. MD5 fails; the application's own non-security MD5 uses (report storage paths, the GraphQL hashedPath fields and avatar URLs) are marked as such so they keep working, and any other MD5 use raises.

Outside that boundary: grpcio bundles BoringSSL and is used only by the Google Cloud Pub/Sub publisher when that integration is configured; pycryptodome ships its own primitives as a dependency of the minio client; argon2-cffi provides the Argon2 password hasher, which is not the default (PBKDF2 through hashlib is); rsa is google-auth's pure-Python fallback; polars carries its own Rust TLS stack, is not imported by the API and is absent from the alpine variants.

Non-hardened images vs. Docker Hardened Images

• The application is installed from the dhi/pkg-codecov-api package at /usr/share/codecov-api with its Python environment at /usr/lib/codecov-api. Upstream copies the whole umbrella monorepo to /app and runs from /app/apps/codecov-api; that path is kept as a symlink, and python and gunicorn on PATH resolve to the packaged environment.
• The image runs as uid 65532 with HOME=/home/nonroot; upstream runs as the codecov user (uid 1001). Bind-mounted configuration must be readable by uid 65532.
• Only the codecov-api dependency group of the upstream lockfile is installed. Worker-only packages upstream's shared requirements image carries (lxml, openai, test-results-parser, ...) are absent, as are the psycopg2-binary and tlslite-ng copies upstream adds under external_deps. The locked psycopg2-binary is built from its source distribution against the system libpq instead of the bundled wheel, and cryptography is built from source against the system OpenSSL instead of the wheel's bundled copy, so PostgreSQL connections and the cryptography primitives use the system libraries.
• The alpine variants install the same locked set without polars, which publishes no musl wheel and which the API never imports; codecov-ribs, which publishes no musl wheel either, is built there from its source distribution.
• The packaged entrypoint carries a one-line fix to the upstream shutdown loop so SIGTERM drains gunicorn for SHUTDOWN_TIMEOUT seconds before it is killed. migrate.sh and migrate-timeseries.sh are executable, as in the upstream image, and the work directory is the upstream path /app/apps/codecov-api, which resolves to /usr/share/codecov-api.
• The runtime image ships bash and coreutils because the upstream entrypoint is a bash script; it ships no other shell tooling and no package manager.

Image variants

Docker Hardened Images come in different variants depending on their intended use. Image variants are identified by their tag.

• Runtime variants are designed to run your application in production. These images are intended to be used either directly or as the FROM image in the final stage of a multi-stage build. These images typically:

  • Run as a nonroot user
  • Do not include a shell or a package manager
  • Contain only the minimal set of libraries needed to run the app

• Build-time variants typically include dev in the tag name and are intended for use in the first stage of a multi-stage Dockerfile. These images typically:

  • Run as the root user
  • Include a shell and package manager
  • Are used to build or compile applications

• FIPS variants include fips in the variant name and tag. They come in both runtime and build-time variants. These variants use cryptographic modules that have been validated under FIPS 140, a U.S. government standard for secure cryptographic operations. For example, usage of MD5 fails in FIPS variants.

To view the image variants and get more information about them, select the Tags tab for this repository, and then select a tag.

Migrate to a Docker Hardened Image

To migrate your application to a Docker Hardened Image, you must update your Dockerfile. At minimum, you must update the base image in your existing Dockerfile to a Docker Hardened Image. This and a few other common changes are listed in the following table of migration notes.

The following steps outline the general migration process.

1. Find hardened images for your app.

   A hardened image may have several variants. Inspect the image tags and find the image variant that meets your needs.

2. Update the base image in your Dockerfile.

   Update the base image in your application's Dockerfile to the hardened image you found in the previous step. For framework images, this is typically going to be an image tagged as dev because it has the tools needed to install packages and dependencies.

3. For multi-stage Dockerfiles, update the runtime image in your Dockerfile.

   To ensure that your final image is as minimal as possible, you should use a multi-stage build. All stages in your Dockerfile should use a hardened image. While intermediary stages will typically use images tagged as dev, your final runtime stage should use a non-dev image variant.

4. Install additional packages

   Docker Hardened Images contain minimal packages in order to reduce the potential attack surface. You may need to install additional packages in your Dockerfile. Inspect the image variants to identify which packages are already installed.

   Only images tagged as dev typically have package managers. You should use a multi-stage Dockerfile to install the packages. Install the packages in the build stage that uses a dev image. Then, if needed, copy any necessary artifacts to the runtime stage that uses a non-dev image.

   For Alpine-based images, you can use apk to install packages. For Debian-based images, you can use apt-get to install packages.

Troubleshooting migration

The following are common issues that you may encounter during migration.

General debugging

The hardened images intended for runtime don't contain a shell nor any tools for debugging. The recommended method for debugging applications built with Docker Hardened Images is to use Docker Debug⁠ to attach to these containers. Docker Debug provides a shell, common debugging tools, and lets you install other tools in an ephemeral, writable layer that only exists during the debugging session.

Permissions

By default image variants intended for runtime, run as the nonroot user. Ensure that necessary files and directories are accessible to the nonroot user. You may need to copy files to different directories or change permissions so your application running as the nonroot user can access them.

Privileged ports

Non-dev hardened images run as a nonroot user by default. As a result, applications in these images can't bind to privileged ports (below 1024) when running in Kubernetes or in Docker Engine versions older than 20.10. To avoid issues, configure your application to listen on port 1025 or higher inside the container, even if you map it to a lower port on the host. For example, docker run -p 80:8080 my-image will work because the port inside the container is 8080, and docker run -p 80:81 my-image won't work because the port inside the container is 81.

No shell

By default, image variants intended for runtime don't contain a shell. Use dev images in build stages to run shell commands and then copy any necessary artifacts into the runtime stage. In addition, use Docker Debug to debug containers with no shell.

Entry point

Docker Hardened Images may have different entry points than images such as Docker Official Images. Use docker inspect to inspect entry points for Docker Hardened Images and update your Dockerfile if necessary.