Performance Optimization

⚡ Performance Optimization

Complete guide for optimizing ZPL2PDF performance and resource usage.

🎯 Overview

Optimize ZPL2PDF for:

• ✅ Processing speed - Faster conversions
• ✅ Memory usage - Reduced RAM consumption
• ✅ Disk I/O - Efficient file operations
• ✅ CPU utilization - Better resource management
• ✅ Scalability - Handle large volumes

---

🚀 Quick Wins

1. Reduce DPI for Faster Processing

# Standard quality (fastest)
ZPL2PDF -i label.txt -o output -d 203

# Low quality (even faster, for testing)
ZPL2PDF -i label.txt -o output -d 150

# Only use high DPI when necessary
ZPL2PDF -i label.txt -o output -d 300  # Slower

2. Use Appropriate Dimensions

# Let ZPL2PDF extract dimensions (faster)
ZPL2PDF -i label.txt -o output

# Only specify when necessary
ZPL2PDF -i label.txt -o output -w 7.5 -h 15 -u in

3. Process Files Sequentially

# Avoid parallel processing (prevents memory spikes)
for file in *.txt; do
    ZPL2PDF -i "$file" -o output
done

---

💾 Memory Optimization

Configuration Settings

{
  "processing": {
    "batchSize": 10,
    "parallelProcessing": false,
    "memoryLimit": "512MB",
    "timeoutSeconds": 300
  }
}

Best Practices

# Process large files individually
ZPL2PDF -i large_file.txt -o output

# Split large files before processing
split -l 100 large_file.txt chunk_

# Clear processed files
rm processed/*.txt  # Manual cleanup

Monitor Memory Usage

# Linux - Monitor memory
top -p $(pgrep ZPL2PDF)
htop -p $(pgrep ZPL2PDF)

# macOS - Monitor memory
top -pid $(pgrep ZPL2PDF)

# Windows - Monitor memory
tasklist /FI "IMAGENAME eq ZPL2PDF.exe"

---

🔧 CPU Optimization

Single vs. Parallel Processing

# Single-threaded (more stable)
ZPL2PDF start -l ./watch

# Process files one at a time
for file in *.txt; do
    ZPL2PDF -i "$file" -o output &
    wait  # Wait for completion before next
done

CPU Affinity (Linux)

# Run on specific CPU cores
taskset -c 0-3 ZPL2PDF start -l ./watch

# Or use systemd service
[Service]
CPUAffinity=0-3

Resource Limits

# Limit CPU usage (Linux)
cpulimit -e ZPL2PDF -l 50  # 50% CPU

# Or use nice
nice -n 10 ZPL2PDF start -l ./watch

# Docker resource limits
docker run --cpus="0.5" --memory="512m" brunoleocam/zpl2pdf:latest

---

💽 Disk I/O Optimization

Fast Storage

# Use SSD for watch and output folders
ZPL2PDF start -l /mnt/ssd/watch -w 7.5 -h 15 -u in

# Or use tmpfs (RAM disk) for temporary files
sudo mount -t tmpfs -o size=512M tmpfs /tmp/zpl2pdf
ZPL2PDF start -l /tmp/zpl2pdf/watch

Reduce File Operations

{
  "autoDeleteProcessed": true,   // Delete after processing
  "outputSubfolders": false,     // Disable subfolders
  "retryDelay": 5000,            // Longer retry delay
  "maxRetries": 2                // Fewer retries
}

Batch Processing

# Process files in batches
find ./watch -name "*.txt" | head -n 10 | while read file; do
    ZPL2PDF -i "$file" -o output
done

# Wait between batches
sleep 5

---

🐳 Docker Optimization

Optimized Image

# Use Alpine variant (smaller, faster)
docker pull brunoleocam/zpl2pdf:alpine

# Multi-stage build (custom)
# See Dockerfile in repository

Resource Limits

version: '3.8'
services:
  zpl2pdf:
    image: brunoleocam/zpl2pdf:alpine
    volumes:
      - ./watch:/app/watch
      - ./output:/app/output
    command: start -l /app/watch -w 7.5 -h 15 -u in
    deploy:
      resources:
        limits:
          memory: 512M
          cpus: '0.5'
        reservations:
          memory: 256M
          cpus: '0.25'

Volume Performance

version: '3.8'
services:
  zpl2pdf:
    image: brunoleocam/zpl2pdf:alpine
    volumes:
      # Use cached/delegated for better performance
      - ./watch:/app/watch:cached
      - ./output:/app/output:delegated
    command: start -l /app/watch

---

📊 Monitoring and Profiling

Performance Metrics

# Measure conversion time
time ZPL2PDF -i label.txt -o output -n label.pdf

# Monitor processing rate
watch -n 1 'ls -l output/ | wc -l'

# Track resource usage
htop -p $(pgrep ZPL2PDF)

Logging for Performance

{
  "logLevel": "Info",  // Use Info (not Debug) in production
  "logToFile": false,  // Disable file logging for speed
  "logToConsole": true
}

Profiling

# Linux - Profile with perf
perf record -g ZPL2PDF -i label.txt -o output
perf report

# macOS - Profile with Instruments
instruments -t "Time Profiler" ZPL2PDF -i label.txt -o output

# Windows - Profile with PerfView
PerfView.exe collect ZPL2PDF.exe -i label.txt -o output

---

🔄 Daemon Mode Optimization

Optimized Configuration

{
  "retryDelay": 5000,        // Longer delay = fewer checks
  "maxRetries": 2,           // Fewer retries = faster failure
  "autoDeleteProcessed": true, // Clean up automatically
  "batchSize": 10,           // Process in small batches
  "watchInterval": 1000      // Check folder every 1 second
}

Watch Folder Best Practices

# Use dedicated SSD partition
ZPL2PDF start -l /mnt/ssd/watch

# Limit folder size
# Use cron to clean old files
0 2 * * * find /var/zpl2pdf/watch -mtime +7 -delete

# Separate input/output folders
ZPL2PDF start -l /var/watch -o /var/output

Periodic Restart

# Restart daemon daily to free memory
# Add to crontab
0 2 * * * ZPL2PDF stop && sleep 5 && ZPL2PDF start -l /var/watch

# Or use systemd timer (Linux)
[Service]
Restart=always
RestartSec=86400  # 24 hours

---

📈 Scaling Strategies

Horizontal Scaling

# Multiple instances for different label sizes
version: '3.8'
services:
  zpl2pdf-small:
    image: brunoleocam/zpl2pdf:alpine
    volumes:
      - ./watch-small:/app/watch
      - ./output:/app/output
    command: start -l /app/watch -w 5 -h 10 -u cm
    
  zpl2pdf-medium:
    image: brunoleocam/zpl2pdf:alpine
    volumes:
      - ./watch-medium:/app/watch
      - ./output:/app/output
    command: start -l /app/watch -w 10 -h 15 -u cm
    
  zpl2pdf-large:
    image: brunoleocam/zpl2pdf:alpine
    volumes:
      - ./watch-large:/app/watch
      - ./output:/app/output
    command: start -l /app/watch -w 15 -h 20 -u cm

Load Balancing

# Use multiple watch folders
mkdir -p watch/{1,2,3,4}

# Start multiple daemons
ZPL2PDF start -l watch/1 -w 7.5 -h 15 -u in &
ZPL2PDF start -l watch/2 -w 7.5 -h 15 -u in &
ZPL2PDF start -l watch/3 -w 7.5 -h 15 -u in &
ZPL2PDF start -l watch/4 -w 7.5 -h 15 -u in &

# Distribute files across folders
# Use custom script or load balancer

---

🧪 Benchmarking

Test Scenarios

# Scenario 1: Single file conversion
time ZPL2PDF -i label.txt -o output -d 203

# Scenario 2: Batch conversion (100 files)
time for i in {1..100}; do
    ZPL2PDF -i "label_${i}.txt" -o output
done

# Scenario 3: Different DPI settings
time ZPL2PDF -i label.txt -o output -d 150  # Low
time ZPL2PDF -i label.txt -o output -d 203  # Standard
time ZPL2PDF -i label.txt -o output -d 300  # High

Performance Comparison

DPI	Processing Time	File Size	Quality
150	0.5s	50KB	Low
203	1.0s	80KB	Standard
300	2.5s	150KB	High
600	6.0s	400KB	Premium

---

💡 Best Practices Summary

Do ✅

• Use standard DPI (203) for most cases
• Process files sequentially
• Monitor resource usage
• Clean up processed files
• Use SSD for watch folders
• Restart daemon periodically
• Set appropriate resource limits

Don't ❌

• Don't use high DPI unnecessarily
• Don't process large batches in parallel
• Don't keep unlimited log files
• Don't use network drives for watch folders
• Don't run without resource limits
• Don't ignore memory warnings

---

🔗 Related Topics

• Configuration - Performance-related settings
• Daemon Mode - Daemon optimization
• Docker Deployment - Container optimization
• Troubleshooting - Performance issues