Skip to main content

Making your CCoE happy

Project description

mansa-ci-pipelinePython VersionsRuffPoetry

Mansa - a FinOps friendly linter

FinOps is an operational framework and cultural practice that enhances the business value derived from cloud technology. It promotes data-driven decision-making in a timely manner and fosters financial responsibility. It gets achieved through collaboration among engineering, finance, and business teams.

FinOps is a discipline that combines financial management principles with cloud engineering and operations to provide organizations with a better understanding of their cloud spending. It also helps them make informed decisions on how to allocate and manage their cloud costs. The goal of FinOps isn't to save money, but to maximize revenue or business value through the cloud. It helps to enable organizations to control cloud spending while maintaining the level of performance, reliability, and security needed to support their business operations.

All of the cloud service providers also allow for the use of tags which can be applied as metadata to most individual resources, and which then appear in the detailed billing reports providing cost and usage data when turned on by the user. Tagging is an essential pre-requirement for FinOps implementation.

Mansa is a lintin tool used internally at B2-Impact to enforce cost-tagging at build time as part of our standard CI-CD pipelines for Azure Machine Learning Resources.

Why Mansa?

Usually IOps teams deploy their infrastructure using terraform/bicep/cdk languages. However in most of the cases the IOps teams are only in charge of deploying the infrastructure "container" that an application will use. In the case of Machine Learning this situation is even more common as the recursive experimental nature of AI applications makes quite common the fact that we do not know which deployment configuration will be best during the development phase.

Besides a lot of MLOps processes as creation of ML pipelines, compute clusters, job launching, inference endpoints etc... are invoked via python SDKs and not terraform code.

Because of that we decided to create Mansa. Mansa scans all your python code in search of classes that accept the tags arguments, currently these classes are configurable with a config.toml file as we have not been able to come up with a better method that is as simple as this.

Mansa also looks that the tags contains a certain key-value pairs with allowed values. Currently this is hardcoded in the code but we plan to make this configurable via a section in the config.toml file.

Why this name?

Mansa Musa was the 9th Mansa of the Mali Empire, he was one of the richest people in history. According to some research [1] he was so rich that he was generating inflation wherever he went. His wealth was not measurable and therefore neither his spending. We took inspiration of his story for the naming of this package as we believe that is important to keep track of cost and we do not want to contribute to the cloud cost inflation phenomena.

Usage guide

Just run:

mansa --directory .
mansa --file mypythonfile.py

If you want to use your custom config.toml file you can use:

mansa --directory . --config myfile.toml
mansa --file mypythonfile.py --config myfile.toml

Build Instructions

  1. Install poetry via the oficial installer:

    curl -sSL https://install.python-poetry.org | python3 -
    
  2. Clone this repository with:

    git clone https://github.com/b2impact/mansa.git
    
  3. Ensure poetry is using the correct virtual environment. You can activate the environment manually with:

    poetry shell
    
  4. Install dependencies and build the package:

    poetry install
    
  5. Build the package:

    poetry build
    

Development guidelines

As usual, we stick to TBD, Create your own branch according to the following guidelines:

  • {type_of_branch}/{namefirstlettersurnamefirstlettersurnamesecondletter}/{change_name}

  • Being type of branch:

    • feat (of feature).

    • bugfix/fix (to fix a bug).

    • enh/improvement (for enhancement, usually runtime performance).

Run tests

To run the tests, run the following in the root directory of the project:

poetry run pytest

Or to run multiversion tests with linting included:

poetry run tox

To install tox we recommend using pipx as:

pipx install tox

Introducing dependencies

Please install dependencies via:

poetry add {dependency-name}=={version}

When doing that your pyproject.toml file will be automatically updated as well as the lock files.

References

[1] Goodwin, A. J. H. (1957). The Medieval Empire of Ghana. The South African Archaeological Bulletin, 12(47), 108–112. https://doi.org/10.2307/3886971

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

mansa-0.1.2.tar.gz (8.5 kB view details)

Uploaded Source

Built Distribution

mansa-0.1.2-py3-none-any.whl (8.7 kB view details)

Uploaded Python 3

File details

Details for the file mansa-0.1.2.tar.gz.

File metadata

  • Download URL: mansa-0.1.2.tar.gz
  • Upload date:
  • Size: 8.5 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/5.1.0 CPython/3.12.5

File hashes

Hashes for mansa-0.1.2.tar.gz
Algorithm Hash digest
SHA256 bf45b7d55c790bd4ddb0bc553e0204dbcc4e94c56c47f15495937d18573a79a5
MD5 06f760b153c93bcbe3b965960866a3bd
BLAKE2b-256 421c067a13fbe6d554f4814f7d812fbb9263b8e0ddf34e92a3fb386d5486c60e

See more details on using hashes here.

File details

Details for the file mansa-0.1.2-py3-none-any.whl.

File metadata

  • Download URL: mansa-0.1.2-py3-none-any.whl
  • Upload date:
  • Size: 8.7 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/5.1.0 CPython/3.12.5

File hashes

Hashes for mansa-0.1.2-py3-none-any.whl
Algorithm Hash digest
SHA256 64cc451b5d19b7a9074d2f2b092bea7c4aa54656be638c875d17ce08478e3bd0
MD5 015e8366c40a6b1233e5cff9bf79d4eb
BLAKE2b-256 289fc4b20a7afcaa87fab69dd2f1edd98283de7f2ffbba38291afb17edf93d1e

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page