User Guide

IN-PROGRESS This comprehensive guide plans to cover all aspects of using PG Scheduler.

• Configuration
  • Scheduler Initialization
  • Misfire Grace Time
  • Job Scheduling Options
  • Environment-Based Configuration
  • Best Practices
  • Next Steps

Overview

PG Scheduler is designed to be both simple for basic use cases and powerful for production deployments. This guide will walk you through:

1. Basic Usage - Core concepts and simple job scheduling

2. Periodic Jobs - Using the @periodic decorator effectively

3. Reliability - Understanding deduplication, retries, and error handling

4. Vacuum Policies - Managing job lifecycle and cleanup

5. Configuration - Tuning the scheduler for your needs

6. Troubleshooting - Common issues and solutions

Key Concepts

Jobs vs Periodic Jobs

• Jobs: One-time executions scheduled for a specific time

• Periodic Jobs: Recurring executions that reschedule themselves

Deduplication

PG Scheduler prevents duplicate job executions across multiple worker replicas using:

• Deterministic Job IDs: Based on function signature and parameters

• Database Constraints: PostgreSQL primary keys prevent duplicates

• Window-based Logic: Time-based deduplication for periodic jobs

Priority System

Jobs can have different priorities:

• JobPriority.CRITICAL - Executes before normal jobs

• JobPriority.NORMAL - Default priority

Reliability Features

• Heartbeat Monitoring: Detect crashed workers

• Orphan Recovery: Clean up abandoned jobs

• Graceful Shutdown: Complete active jobs before stopping

• Retry Logic: Automatic retry with exponential backoff