heroku-manager 0.1.2

A package for managing Heroku dynos with autoscaling capabilities

Heroku Manager

A Python package for managing Heroku dynos with autoscaling capabilities.

Features

• Automatic scaling of Heroku dynos based on memory usage and load
• Monitoring of dyno health and performance
• Automatic restart of unresponsive dynos
• Integration with Django for caching and configuration

Installation

pip install heroku-manager

Usage

from heroku_manager import HerokuManager

# Get the autoscaler instance
autoscaler = HerokuManager.get_autoscaler()

# Start continuous autoscaling
autoscaler.start_continuous_autoscale()

# Manual scaling operations
if autoscaler.requires_upscale:
    autoscaler.upscale_formation_to_next_level()
elif autoscaler.allow_downscale:
    autoscaler.downscale_formation_to_original_size()

# Stop autoscaling
autoscaler.stop_continuous_autoscale()

Configuration

The package requires the following environment variables:

• HEROKU_API_KEY: Your Heroku API key
• HEROKU_APP_NAME: The name of your Heroku app
• DYNO: The name of the current dyno (automatically set by Heroku)

Django Settings

When used with Django, the following settings are available:

• DYNO_CONTINUOS_AUTOSCALE_ENABLED: Enable continuous autoscaling
• DYNO_AUTOSCALE_INTERVAL: Interval between autoscale checks (in seconds)
• DYNO_TIME_BETWEEN_SCALES: Minimum time between scaling operations (in seconds)
• DYNO_MIN_UPSCALE_DURATION: Minimum duration to keep a dyno upscaled (in seconds)
• DYNO_DOWNSCALE_CHECK_INTERVAL: Interval for checking if a dyno can be downscaled (in seconds)
• DYNO_ZOMBIE_THRESHOLD: Time threshold for considering a dyno as unresponsive (in seconds)
• DYNO_LOG_THREADS_USED: Whether to log the number of threads used
• DYNO_LOGS_CACHE_DURATION: Duration to cache dyno logs (in seconds)
• DYNO_ERRORS_TIMEOUT_DURATION: Duration to cache error information (in seconds)
• DYNO_GENERAL_CACHE_DURATION: General cache duration (in seconds)
• DYNO_TIME_BETWEEN_RESTARTS: Minimum time between dyno restarts (in seconds)
• DYNO_AUTOSCALE_ENABLED_FOR_BEATWORKER: Whether to enable autoscaling for beat workers
• UPSCALE_PERCENTAGE_HIGH_MEM_USE: Memory usage percentage threshold for upscaling
• DOWNSCALE_PERCENTAGE_HIGH_MEM_USE: Memory usage percentage threshold for downscaling
• HIGH_MEM_USE_MB: High memory usage threshold in MB