laravel/horizon
Horizon
Horizon is a background job orchestrator and worker manager for Redis queues. It provides a monitoring dashboard to track job throughput, wait times, and failure rates, alongside a system for managing job retries, execution timeouts, and worker distribution.
The project distinguishes itself through a Redis-backed monitoring interface that identifies system bottlenecks and a queue alerting system that sends notifications when background job wait times exceed defined thresholds. Worker processes are managed via version-controlled configuration files to ensure consistent balancing and scaling across different environments.
The system covers broad capability areas including performance metrics tracking, workload distribution balancing, and job lifecycle management. It includes administrative utilities for purging failed jobs and a development workflow that automatically restarts workers upon code changes.
Access to the monitoring dashboard is restricted via authorization gates to control visibility in non-local environments.
Features
- Background Job Dispatchers - Acts as a central orchestrator for managing job retries, timeouts, and worker distribution.
- Background Job Processing - Provides a comprehensive system for offloading and managing asynchronous background tasks.
- Work Queue Performance Metrics - Provides detailed metrics on job execution time and volume to identify processing bottlenecks.
- Queue Workload Balancing - Allocates worker processes across queues using workload-based scaling, even distribution, or priority ordering.
- Worker Process Management - Provides a configuration system for defining worker counts and supervisor groups to control background job processing.
- Queue Worker Management - Manages the orchestration and scaling of long-lived worker processes consuming tasks from Redis queues.
- Redis-Backed Queues - Utilizes Redis as the primary storage and broker for managing job states and metadata.
- Queue Monitoring Dashboards - Provides a visual interface to identify system bottlenecks by tracking job throughput and failure rates.
- Retry Strategies - Implements configurable retry logic with maximum attempt counts and exponential backoff strategies.
- Alert Thresholds - Sends notifications via external messaging services when queue wait times exceed predefined numerical thresholds.
- Metric Dashboards - Ships a visual dashboard that tracks throughput, runtime, and failure rates to identify system bottlenecks.
- Metric and Performance Monitors - Provides high-frequency collection and visualization of job throughput and wait time metrics.
- Job Tagging - Implements a system for assigning searchable metadata tags to jobs for improved tracking and grouping.
- Development Iteration Workflows - Speeds up the development cycle by automatically restarting workers upon code changes.
- Execution Timeout Management - Allows specifying the maximum duration a worker can process a job before the process is terminated.
- Task Worker Configurations - Uses version-controlled PHP files to define worker execution parameters, timeouts, and retry logic.
- Worker Scaling - Provides configuration-driven scaling of background worker processes to manage processing capacity.
- Dashboard Access Controls - Restricts visibility of the monitoring dashboard in non-local environments using configurable authorization policies.
- Wait-Time Alerting - Sends notifications when background job backlogs exceed defined time thresholds to maintain responsiveness.
- Queue Metric Polling - Periodically queries queue lengths and job runtimes to generate performance statistics.
- Queue Performance Tuning - Balances worker distribution and configures execution timeouts to optimize background workload processing.
- Worker Live Reloading - Provides a utility to automatically reload worker processes when source code changes are detected.
- Monitoring - Monitors queued jobs.
