Code Monkey home page Code Monkey logo

confsecrets's Introduction

confsecrets

Build Status Coverage Status

A simple utility module to symmetrically encrypt/decrypt application secrets.

Description

It is often difficult for developers to secure passwords for databases, search engines, directory services, etc. Security wishes to make sure these secrets are centralized, but this adds a dependency on an external service, not to mention code complexity.

Several solutions exist, but these are often silos without any abstraction built around them.

The goal of this project is to provide APIs that wrap the simplest solutions that are actually solutions, namely:

  • Keeping passwords as encrypted values on the filesystem, with a passphrase in the code.

Future versions will add the ability to keep the encrypted material in S3, or another cloud storage provider, and to keep the passphrase from which a key is derived also in cloud storage.

Other Tools

pyjks

  • Doesn't support saving jceks
  • Can load from string, so we could use it as keystore format and add S3 and stuff

keyczar

  • Manages encryption, but doesn't support MFA keys - e.g. not a vault in which to store encrypted material.
  • Doesn't directly support use of S3 or HashiCorp vault as a backend, only itself.
  • Doesn't offer management of secrets, just management of keys.

dynaconf

See https://www.dynaconf.com/ - this provides a superset of the features in confsecrets at present, but also appears more complicated to use. I also cannot say dynaconf is up to U.S.Federal Government standards.

EC2 Secrets Manager

  • Again, this is a central play, and locks you in somewhat to the vendor.

Concept of Operation

The goal here is to reach the point where we can keep multiple secrets in the same encrypted wodge, decrypted with the same passwords, and provide some command-line over them that is easily integrated into Django. We want to act locally but think globally.

How about this:

  • implement confsecrets\vault.py which contains a Vault class that is a subclass of UserDict, maybe OrderedDict.

  • The vault has the following init:

      Vault(key=, salt=, path=)
    
  • It encrypts using the the salt and key, converting the crypttext to base64.

  • Encrypting immediately flushes the file that has been loaded

  • On decrypt, it checks to see if the file has changed, and reloads the underlying data in that case.

  • On decrypt, it reverses the process.

There is a concept of a "default vault" whose configuration is controlled by environment variables or through the API. The "default vault" is is a singleton.

Django integration is provided via a confsecrets.django application that allows the configuration to be provided by settings:

CONFSECRETS_SALT = b'89982hto'
CONFSECRETS_KEY = 'This is not an example'
CONFSECRETS_VAULT = os.path.join(BASE_DIR, 'vault.yaml')

This initializes the default vault during configuration freeze. Otherwise, the default vault's configuration is controlled by the environment variables.

This is secure as long as the vault file is not stored in git, and then it becomes obfuscation. Alternatively, the vault file could be stored in the project, but the key is external to the git repository.

Saving secrets becomes easy through a command-line/management command to populate the vault:

  • list - lists the secrets stored in the vault, along with their value
  • write <name> <value> - uses value if present, otherwise uses stdin
  • read <name> - typical options, outputs the secret to the stdout
  • rm <name> - removes an encrypted value from the vault

With the system configured, dealing with the vault becomes as easy as using Secret objects:

ELASTICSEARCH_PASSWORD = Secret('name')

To access it, you can treat it like a string:

from django.conf import settings

settings.ES_PASSWORD.decrypt()

The secret also acts like a string:

int(str(settings.ES_PASSWORD))

or:

auth_string = "%s:%s" % (username, settings.ES_PASSWORD)

This will fail for a number of reasons with clear exceptions:

  • If the configuration is not available via settings or environment variables
  • If the vault is not-present
  • If the vault doesn't have that key in it

Vault would then know how to deal with the operations described above.

Self-critique

This way of doing things does not fully comply the Twelve-Factor App way of managing the environment. Django's settings are similarly response to the environment while recognizing that configuration solely via the environment is complicated.

Roadmap

Not sure on the priority of these:

  • Add support for placing the vault on S3. Vault becomes polymorphic because if path is an URL, then we will create a different sub-class of Vault using an override of __new__. A local path vault is still standard.

  • Separate cli from Django management commands.

  • Add support for placing the passphrase on S3 similarly.

  • Support secret versioning, so that it is possible for a developer to push a new value for a password from the desktop, and then change the password to some backend.

  • Add support for a default value for a secret, so that you get the default instead of KeyError

  • Add support/integration for Flask - it can surely be used without this because pbe, secret, and vault sub-modules are independent.

  • Move to cryptodome once that package is out of Beta.

Status

Have implemented the Vault, DefaultVault, and initial Secret facilities. Django management commands and integration needs tests.

confsecrets's People

Contributors

danizen avatar

Watchers

 avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.