AAP metrics-utility

A standalone CLI utility for collecting and reporting metrics from Ansible Automation Platform (AAP) Controller instances. This tool allows users to:

• Collect and analyze Controller usage data
• Generate reports (CCSP, CCSPv2, RENEWAL_GUIDANCE)
• Support multiple storage adapters for data persistence (local dir, S3)
• Push metrics data to console.redhat.com

Quick Start

The project can run in 2 modes - real controller, and standalone.

The controller mode is for running inside Controller containers and is more fully documented in docs/old-readme.md.

The standalone mode is currently used only for development & testing. It does not need a running awx instance (only a running postgres with imported data), and mocks some values otherwise obtained from awx (see mock_awx/settings/__init__.py).

Prerequisites (standalone)

• Python 3.11 or later
• uv package manager (Install with pip install uv if not already installed)
• Dependencies managed via pyproject.toml (Ensure uv.lock is used for consistency)
• MacOS brew install postgresql, if you need the database or minio

Installation (standalone)

# Clone the repository
git clone https://github.com/ansible/metrics-utility.git
cd metrics-utility

# Install dependencies using uv
uv sync

For more about the development setup, see the Developer Setup Guide.

Tests (standalone)

Run tests using uv run pytest -s -v. Some tests depend on a running postgres & minio instance - run docker compose -f tools/docker/docker-compose.yaml up to get one.

You can also run pytest inside a container too - to run all tests once, you can docker compose -f tools/docker/docker-compose.yaml --profile=pytest up. You use also podman.

For more flexibility, use:

(host) $ docker compose -f tools/docker/docker-compose.yaml --profile=env up -d  # runs a metrics-utility-env container with python & uv set up
(host) $ docker exec -it metrics-utility-env /bin/sh # (wait for postgres & minio containers to start before running)
(container) $ uv run pytest -vv metrics_utility/test/ccspv_reports/test_complex_CCSP_with_scope.py # 1 test
(container) $ uv run pytest -vv metrics_utility/test/ccspv_reports # all ccsp tests

Using Docker (in CI mode to be able to run all tests)

# Ensure SQL data is loaded (only needed once after starting containers)
docker compose -f tools/docker/docker-compose.yaml exec postgres bash -c \
  'cat /docker-entrypoint-initdb.d/init-*.sql | psql -U awx -d postgres'

# Run all gather tests
docker compose -f tools/docker/docker-compose.yaml exec metrics-utility-env bash -c \
  'sed -i "/NAME/s/awx/postgres/" mock_awx/settings/__init__.py && \
   sed -i "/USER/s/myuser/awx/" mock_awx/settings/__init__.py && \
   sed -i "/PASSWORD/s/mypassword/awx/" mock_awx/settings/__init__.py && \
   sed -i "/HOST.*localhost/s/localhost/postgres/" mock_awx/settings/__init__.py && \
   uv run pytest -s -v metrics_utility/test/gather/'

# Run a specific gather test
docker compose -f tools/docker/docker-compose.yaml exec metrics-utility-env bash -c \
  'sed -i "/NAME/s/awx/postgres/" mock_awx/settings/__init__.py && \
   sed -i "/USER/s/myuser/awx/" mock_awx/settings/__init__.py && \
   sed -i "/PASSWORD/s/mypassword/awx/" mock_awx/settings/__init__.py && \
   sed -i "/HOST.*localhost/s/localhost/postgres/" mock_awx/settings/__init__.py && \
   uv run pytest -s -v metrics_utility/test/gather/test_jobhostsummary_gather.py::test_command'

# Run all tests (not just gather)
docker compose -f tools/docker/docker-compose.yaml exec metrics-utility-env bash -c \
  'sed -i "/NAME/s/awx/postgres/" mock_awx/settings/__init__.py && \
   sed -i "/USER/s/myuser/awx/" mock_awx/settings/__init__.py && \
   sed -i "/PASSWORD/s/mypassword/awx/" mock_awx/settings/__init__.py && \
   sed -i "/HOST.*localhost/s/localhost/postgres/" mock_awx/settings/__init__.py && \
   uv run pytest -s -v'

Using Podman (in CI mode to be able to run all tests)

# Ensure SQL data is loaded (only needed once after starting containers)
podman compose -f tools/docker/docker-compose.yaml exec postgres bash -c \
  'cat /docker-entrypoint-initdb.d/init-*.sql | psql -U awx -d postgres'

# Run all gather tests
podman compose -f tools/docker/docker-compose.yaml exec metrics-utility-env bash -c \
  'sed -i "/NAME/s/awx/postgres/" mock_awx/settings/__init__.py && \
   sed -i "/USER/s/myuser/awx/" mock_awx/settings/__init__.py && \
   sed -i "/PASSWORD/s/mypassword/awx/" mock_awx/settings/__init__.py && \
   sed -i "/HOST.*localhost/s/localhost/postgres/" mock_awx/settings/__init__.py && \
   uv run pytest -s -v metrics_utility/test/gather/'

# Run a specific gather test
podman compose -f tools/docker/docker-compose.yaml exec metrics-utility-env bash -c \
  'sed -i "/NAME/s/awx/postgres/" mock_awx/settings/__init__.py && \
   sed -i "/USER/s/myuser/awx/" mock_awx/settings/__init__.py && \
   sed -i "/PASSWORD/s/mypassword/awx/" mock_awx/settings/__init__.py && \
   sed -i "/HOST.*localhost/s/localhost/postgres/" mock_awx/settings/__init__.py && \
   uv run pytest -s -v metrics_utility/test/gather/test_jobhostsummary_gather.py::test_command'

# Run all tests (not just gather)
podman compose -f tools/docker/docker-compose.yaml exec metrics-utility-env bash -c \
  'sed -i "/NAME/s/awx/postgres/" mock_awx/settings/__init__.py && \
   sed -i "/USER/s/myuser/awx/" mock_awx/settings/__init__.py && \
   sed -i "/PASSWORD/s/mypassword/awx/" mock_awx/settings/__init__.py && \
   sed -i "/HOST.*localhost/s/localhost/postgres/" mock_awx/settings/__init__.py && \
   uv run pytest -s -v'

Basic Usage

1. Know the environment

• In Controller mode:
  • make sure to connect to a running Controller instance,
  • get metrics-utility (map a volume, or git clone),
  • activate the virtual environment (source /var/lib/awx/venv/awx/bin/activate),
  • pip install . from the metrics-utility dir,
  • run utility using python manage.py ....
• In RPM mode:
  • install the right RPM
  • run utility using metrics-utility ....
• In standalone mode:
  • make sure to run docker compose -f tools/docker/docker-compose.yaml up if you need the database or minio,
  • run utility using uv run python manage.py ....

1. Pick a task (goes right after the previous command)

• gather_automation_controller_billing_data - collects metrics from controller db, saves daily tarballs with csv/json inside
• build_report - builds a XLSX report, either from controller db or collected tarballs

1. Pick a report type (export METRICS_UTILITY_REPORT_TYPE=...)

• CCSPv2 - uses metrics tarballs to produce a usage report
• CCSP - similar to v2, slightly different aggregation
• RENEWAL_GUIDANCE - uses controller db to produce a renewal guidance report

1. Pick a time period

• --since=12m
• and --until=10m (only with CCSP and CCSPv2)
• or --month=2024-06 (only with build_report)

1. Use --help to see any other params

• build_report also supports --ephemeral, --force and --verbose
• gather_automation_controller_billing_data also supports --dry-run and --ship

1. Set any other necessary environmental variable - see more in docs/old-readme.md.

• METRICS_UTILITY_BILLING_ACCOUNT_ID
• METRICS_UTILITY_BILLING_PROVIDER
• METRICS_UTILITY_BUCKET_ACCESS_KEY
• METRICS_UTILITY_BUCKET_ENDPOINT
• METRICS_UTILITY_BUCKET_NAME
• METRICS_UTILITY_BUCKET_REGION
• METRICS_UTILITY_BUCKET_SECRET_KEY
• METRICS_UTILITY_CRC_INGRESS_URL
• METRICS_UTILITY_CRC_SSO_URL
• METRICS_UTILITY_OPTIONAL_CCSP_REPORT_SHEETS
• METRICS_UTILITY_OPTIONAL_COLLECTORS
• METRICS_UTILITY_ORGANIZATION_FILTER
• METRICS_UTILITY_PRICE_PER_NODE
• METRICS_UTILITY_PROXY_URL
• METRICS_UTILITY_RED_HAT_ORG_ID
• METRICS_UTILITY_REPORT_COMPANY_BUSINESS_LEADER
• METRICS_UTILITY_REPORT_COMPANY_NAME
• METRICS_UTILITY_REPORT_COMPANY_PROCUREMENT_LEADER
• METRICS_UTILITY_REPORT_EMAIL
• METRICS_UTILITY_REPORT_END_USER_CITY
• METRICS_UTILITY_REPORT_END_USER_COMPANY_NAME
• METRICS_UTILITY_REPORT_END_USER_COUNTRY
• METRICS_UTILITY_REPORT_END_USER_STATE
• METRICS_UTILITY_REPORT_H1_HEADING
• METRICS_UTILITY_REPORT_PO_NUMBER
• METRICS_UTILITY_REPORT_RHN_LOGIN
• METRICS_UTILITY_REPORT_SKU
• METRICS_UTILITY_REPORT_SKU_DESCRIPTION
• METRICS_UTILITY_REPORT_TYPE
• METRICS_UTILITY_SERVICE_ACCOUNT_ID
• METRICS_UTILITY_SERVICE_ACCOUNT_SECRET
• METRICS_UTILITY_SHIP_PATH
• METRICS_UTILITY_SHIP_TARGET

Example CCSPv2 run

# You can use also an env-file but then you must export it with `export UV_ENV_FILE=<your_env_file>`
export METRICS_UTILITY_REPORT_TYPE="CCSPv2"
export METRICS_UTILITY_SHIP_PATH="./test/test_data/"
export METRICS_UTILITY_SHIP_TARGET="directory"

export METRICS_UTILITY_PRICE_PER_NODE=11.55 # in USD
export METRICS_UTILITY_REPORT_COMPANY_NAME="Partner A"
export METRICS_UTILITY_REPORT_EMAIL="email@email.com"
export METRICS_UTILITY_REPORT_END_USER_CITY="Springfield"
export METRICS_UTILITY_REPORT_END_USER_COMPANY_NAME="Customer A"
export METRICS_UTILITY_REPORT_END_USER_COUNTRY="US"
export METRICS_UTILITY_REPORT_END_USER_STATE="TX"
export METRICS_UTILITY_REPORT_H1_HEADING="CCSP NA Direct Reporting Template"
export METRICS_UTILITY_REPORT_PO_NUMBER="123"
export METRICS_UTILITY_REPORT_RHN_LOGIN="test_login"
export METRICS_UTILITY_REPORT_SKU="MCT3752MO"
export METRICS_UTILITY_REPORT_SKU_DESCRIPTION="EX: Red Hat Ansible Automation Platform, Full Support (1 Managed Node, Dedicated, Monthly)"

# collect data
uv run ./manage.py gather_automation_controller_billing_data --ship --until=10m --force

# collected tarballs somewhere here (by date and instance uuid)
ls metrics_utility/test/test_data/data/2024/04/*

# build report, overwrite existing if necessary
uv run ./manage.py build_report --month=2024-04 --force

# resulting XLSX
ls metrics_utility/test/test_data/reports/2024/04/

Example RENEWAL_GUIDANCE run

export METRICS_UTILITY_REPORT_TYPE="RENEWAL_GUIDANCE"
export METRICS_UTILITY_SHIP_PATH="./out"
export METRICS_UTILITY_SHIP_TARGET="controller_db"

uv run ./manage.py build_report --since=12months --ephemeral=1month --force

Documentation

Documentation is available in the /docs directory.

Please note that this is the upstream documentation for the metrics-utility project. Additional internal downstream documentation, accessible only to the Ansible organization, is maintained separately in the Ansible Handbook.

As the project grows, more guides and references will be added to the /docs folder.

Version mapping

Contributing

Please follow our Contributor's Guide for details on submitting changes and documentation standards.