Skip to main content

​
Scheduler Deployment

Deploy scheduled agent and recipe execution for 24/7 autonomous operations in production environments.

​
Overview

The scheduler provides:
  • Interval-based agent/recipe execution
  • PM2-style daemon management
  • Cost budgeting and monitoring
  • Automatic retry with exponential backoff
  • Centralized logging

​
Quick Start

​
Start a Scheduler

# With a task prompt
praisonai schedule start news-bot "Check AI news" --interval hourly

# With a recipe
praisonai schedule start news-monitor --recipe news-analyzer --interval hourly

# With all options
praisonai schedule start my-scheduler \
    --recipe my-recipe \
    --interval "*/6h" \
    --timeout 600 \
    --max-cost 2.00 \
    --max-retries 3

​
Manage Schedulers

# List all schedulers
praisonai schedule list

# View logs
praisonai schedule logs news-bot --follow

# Stop scheduler
praisonai schedule stop news-bot

# Restart scheduler
praisonai schedule restart news-bot

​
Python Deployment

from praisonai import recipe

# Schedule a recipe
scheduler = recipe.schedule(
    "my-recipe",
    interval="hourly",
    max_retries=3,
    timeout_sec=300,
    max_cost_usd=1.00,
    run_immediately=True,
)

# Start the scheduler
scheduler.start()

# Get statistics
stats = scheduler.get_stats()
print(f"Executions: {stats['total_executions']}")
print(f"Success rate: {stats['success_rate']}%")

# Stop when done
scheduler.stop()

​
Docker Deployment

​
Dockerfile

FROM python:3.11-slim

WORKDIR /app

# Install dependencies
COPY requirements.txt .
RUN pip install -r requirements.txt

# Copy application
COPY . .

# Set environment variables
ENV OPENAI_API_KEY=${OPENAI_API_KEY}

# Run scheduler
CMD ["praisonai", "schedule", "start", "my-scheduler", "--recipe", "my-recipe", "--interval", "hourly"]

​
Docker Compose

version: '3.8'

services:
  scheduler:
    build: .
    environment:
      - OPENAI_API_KEY=${OPENAI_API_KEY}
    volumes:
      - scheduler-data:/root/.praisonai
    restart: unless-stopped

volumes:
  scheduler-data:

​
Configuration

​
Schedule Intervals

FormatIntervalDescription
hourly3600sEvery hour
daily86400sEvery 24 hours
*/30m1800sEvery 30 minutes
*/6h21600sEvery 6 hours
*/5s5sEvery 5 seconds (testing)
36003600sCustom seconds

​
TEMPLATE.yaml Runtime Block

Configure scheduler defaults in your recipe: 
runtime:
  schedule:
    enabled: true
    interval: "hourly"
    max_retries: 3
    run_immediately: false
    timeout_sec: 300
    max_cost_usd: 1.00

​
Safe Defaults

SettingDefaultDescription
intervalhourlyExecution interval
max_retries3Retry attempts on failure
timeout_sec300Timeout per execution
max_cost_usd1.00Budget limit

​
Production Considerations

​
Cost Monitoring

Set budget limits to prevent runaway costs: 
praisonai schedule start my-scheduler \
    --recipe my-recipe \
    --interval hourly \
    --max-cost 10.00
The scheduler automatically stops when the budget is reached: 
Budget limit reached: $10.01 >= $10.00
Stopping scheduler to prevent additional costs

​
Logging

Logs are stored in ~/.praisonai/logs/: 
# View logs
praisonai schedule logs my-scheduler

# Follow logs in real-time
praisonai schedule logs my-scheduler --follow

# Logs location
ls ~/.praisonai/logs/

​
State Persistence

Scheduler state is persisted in ~/.praisonai/schedulers/: 
# View state files
ls ~/.praisonai/schedulers/

# Export scheduler config
praisonai schedule save my-scheduler --output config.yaml

​
Error Handling

The scheduler automatically retries failed executions with exponential backoff:
  • Attempt 1: Execute immediately
  • Attempt 2: Wait 30s, retry
  • Attempt 3: Wait 60s, retry
  • Attempt 4: Wait 90s, retry
  • Attempt 5: Wait 120s, retry

​
Monitoring

# Show scheduler statistics
praisonai schedule stats my-scheduler

# Describe scheduler details
praisonai schedule describe my-scheduler
Output: 
Scheduler: my-scheduler
========================
Status: running
PID: 12345
Recipe: my-recipe
Interval: hourly
Timeout: 300s
Budget: $10.00

Statistics:
  Total Executions: 24
  Successful: 23
  Failed: 1
  Success Rate: 95.8%
  Total Cost: $0.0024

​
Kubernetes Deployment

apiVersion: apps/v1
kind: Deployment
metadata:
  name: praisonai-scheduler
spec:
  replicas: 1  # Only one replica for scheduler
  selector:
    matchLabels:
      app: praisonai-scheduler
  template:
    metadata:
      labels:
        app: praisonai-scheduler
    spec:
      containers:
      - name: scheduler
        image: praisonai/scheduler:latest
        command: ["praisonai", "schedule", "start", "my-scheduler", "--recipe", "my-recipe", "--interval", "hourly"]
        env:
        - name: OPENAI_API_KEY
          valueFrom:
            secretKeyRef:
              name: praisonai-secrets
              key: openai-api-key
        volumeMounts:
        - name: scheduler-data
          mountPath: /root/.praisonai
        resources:
          requests:
            memory: "256Mi"
            cpu: "100m"
          limits:
            memory: "1Gi"
            cpu: "500m"
      volumes:
      - name: scheduler-data
        persistentVolumeClaim:
          claimName: scheduler-pvc
---
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: scheduler-pvc
spec:
  accessModes:
    - ReadWriteOnce
  resources:
    requests:
      storage: 1Gi

​
Systemd Service

For Linux deployments, create a systemd service: 
# /etc/systemd/system/praisonai-scheduler.service
[Unit]
Description=PraisonAI Scheduler
After=network.target

[Service]
Type=simple
User=praisonai
Environment=OPENAI_API_KEY=your-key-here
ExecStart=/usr/local/bin/praisonai schedule start my-scheduler --recipe my-recipe --interval hourly
Restart=always
RestartSec=10

[Install]
WantedBy=multi-user.target
Enable and start: 
sudo systemctl enable praisonai-scheduler
sudo systemctl start praisonai-scheduler
sudo systemctl status praisonai-scheduler

​
See Also