Django Simple Secrets

A Django integration for AWS Secrets Manager with caching and lazy loading support.

Overview

django-simple-secrets provides a clean and efficient way to access AWS Secrets Manager from a Django applications. It features built-in caching to reduce API calls and supports lazy loading of secrets through Django's SimpleLazyObject.

⚠️ Important Note: This module is not thread-safe. The caching mechanism uses a simple dictionary that is not protected against concurrent access. If you need thread-safety, consider implementing your own synchronization mechanism or disabling caching with use_cache=False.

Key Features

• Simple, intuitive API for accessing AWS Secrets Manager
• Built-in caching mechanism to minimize API calls
• Lazy loading support for improved performance
• Configured retry logic and timeouts
• Type hints for better IDE support
• Compatible with Python 3.10+ and Django 4.x/5.x

Installation

pip install django-simple-secrets

Note: This package requires boto3, which must be installed separately:

pip install boto3

Configuration

Ensure your AWS credentials are properly configured either through:

• Environment variables (AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY)
• AWS credentials file
• IAM role (when running on AWS infrastructure)

Usage

Basic Usage

from django_secrets import get_secret

# Get an entire secret
database_config = get_secret('prod/database')

# Get a specific key from a secret
database_password = get_secret('prod/database', key='password')

Lazy Loading

from django_secrets import get_secret_lazy

# Secret won't be fetched until actually accessed
database_config = get_secret_lazy('prod/database')

Cache Management

from django_secrets import clear_cache

# Clear a specific secret from cache
clear_cache('prod/database')

# Clear entire cache
clear_cache()

# Bypass cache for a single request
fresh_secret = get_secret('prod/database', use_cache=False)

Error Handling

The module raises standard boto3 exceptions:

try:
    secret = get_secret('prod/database')
except ClientError as e:
    # Handle AWS-specific errors
    pass
except KeyError as e:
    # Handle missing key errors
    pass

Performance Considerations

• The module maintains a single boto3 client instance
• Caching is enabled by default to minimize API calls
• Timeout settings: 2 seconds for both connect and read
• Maximum retry attempts: 3

License

This project is licensed under the MIT License. See the LICENSE file for details.