# Worker

The QuickDapp worker system provides simple background job processing for maintenance tasks. It uses child processes and a database-backed queue for basic job scheduling and execution.

# Architecture

The worker system is simple and focused:

WorkerManager - Manages worker processes with basic start/stop functionality
Job Queue - Database storage using the workerJobs table
Job Types - Three built-in job types for system maintenance
Scheduling - Cron-based scheduling with automatic rescheduling

# Key Features

# Simple Process Model

Workers run as child processes:

Basic process isolation for job execution
Configurable worker count (including auto-scaling with 'cpus')
Graceful shutdown support

# Database Storage

Jobs are stored in the workerJobs table:

Persistence - Jobs survive server restarts
Scheduling - Time-based execution with due timestamps
Status Tracking - Simple success/failure tracking
Cleanup - Automatic removal of old jobs

# Built-in Job Types

QuickDapp includes three system maintenance jobs:

# removeOldWorkerJobs

Cleans up old completed worker jobs:

// Scheduled maintenance job
// Runs periodically to clean up old completed jobs

# watchChain

Monitors blockchain events and processes new transactions:

// Blockchain monitoring job
// Watches for contract events and processes them

# deployMulticall3

Deploys the Multicall3 contract if not present:

// Contract deployment job
// Ensures Multicall3 is available on the current chain

# Worker Configuration

Configure workers through environment variables:

WORKER_COUNT=cpus        # Number of worker processes ('cpus' for auto-scale)
WORKER_LOG_LEVEL=info    # Worker-specific log level

# Basic Usage

The worker system is automatically started with the server and runs maintenance tasks in the background. Jobs are scheduled automatically based on cron expressions and system needs.

# WorkerManager Interface

The WorkerManager provides a simple interface:

interface WorkerManager {
  submitJob(job: WorkerJob): Promise<WorkerJob>
  getWorkerCount(): number
  shutdown(): Promise<void>
}

# Job Structure

Jobs in the database have this structure:

interface WorkerJob {
  id: number
  type: 'removeOldWorkerJobs' | 'watchChain' | 'deployMulticall3'
  userId: number
  data: any
  due: Date
  started?: Date
  finished?: Date
  success?: boolean
  result?: any
  cronSchedule?: string
  persistent: boolean
}

# Monitoring

The worker system provides basic status information through health endpoints and logs. Workers automatically handle failures and reschedule jobs when appropriate.

For more details on adding custom jobs, see Adding Jobs.