The Application Autoscaler for Cloud Foundry is a release repository that enables automated scaling of applications running on the Cloud Foundry platform. Written primarily in Shell, the project provides three distinct scaling mechanisms: dynamic scaling based on application performance metrics, dynamic scaling based on custom metrics, and scheduled scaling based on time. This multi-faceted approach allows Cloud Foundry operators to adjust computation resources for applications automatically without manual intervention.
The repository is classified across multiple domains including cloud foundry, containerization, scaling policies, capacity planning, dynamic scaling, cloud-native resource management, kubernetes, microservices, load balancing, and performance optimization. These classifications reflect the breadth of the autoscaler's functionality in managing application workloads across distributed systems.
Development on the project is facilitated through a comprehensive local development setup that leverages Docker for database management, devbox for tool provisioning, and direnv for automatic environment configuration. The repository includes an extensive set of make targets supporting various testing scenarios. Developers can run unit tests and integration tests against multiple database backends including PostgreSQL and MySQL, with support for testing against specific database versions. Additional targets enable acceptance testing, code linting with optional auto-fixes, project compilation, and deployment of the autoscaler with service broker registration in Cloud Foundry environments.
The autoscaler architecture comprises multiple microservices, each with dedicated health monitoring endpoints. These components include the eventgenerator, metricsforwarder, scalingengine, operator, and scheduler, all protected with basic authentication credentials managed through credhub. Health endpoints can be disabled through provided operations files for environments where basic auth protection is unnecessary.
Activity tracking shows that the project maintains active issue and pull request management, with a median response latency of 15.8 hours across tracked items. Bug reports represent the most frequently labeled issue type with six instances, followed by backlog items with three entries and enhancements with one. The project's core contributors include silvestre with twelve tracked events, bonzofenix with six events, and damzog with two events, indicating a relatively focused contributor base. The repository shares overlapping contributors with other Cloud Foundry projects including bosh, cli, and uaa, suggesting integration within the broader Cloud Foundry ecosystem.
The project provides comprehensive documentation through user guides covering policy definition, supported metrics, and public API specifications, along with command-line tooling for interacting with the autoscaler service. The codebase is released under Apache License 2.0, making it available for both open-source and commercial use within the Cloud Foundry community.
