Memory Settings Guide

Optimize BrowserStack Code Quality memory allocation for efficient analysis of large codebases.

Two Analysis Modes — Where to Configure

BrowserStack Code Quality supports two distinct analysis modes, and memory must be tuned at the right place for each:

Analysis Mode	Where Analysis Runs	Where to Set Memory
Server (UI / Docker)	Inside the Docker container on the server	Docker environment variables (-e ANALYSER_XMX=...)
Remote / CLI	On the machine running the embold-scanner CLI (your laptop or CI/CD agent)	Shell environment variables on that machine (e.g., export ANALYSER_XMX=...)

The environment variable names are the same in both modes — the difference is only where you set them.

---

Overview

For repositories with 2+ million lines of code, proper memory configuration is essential for successful analysis. Two key settings control memory usage:

• Container / Process Memory: Total memory available to the analysis process
• Analyzer Heap (ANALYSER_XMX): Memory specifically for the analysis engine (JVM heap)

Memory Recommendations

Size-Based Configuration

Lines of Code	ANALYSER_XMX	Container Memory (-m)	Recommended RAM
Up to 1 Million	-Xmx6g	12GB	16 GB
2-10 Million	-Xmx15g	30GB	32 GB
10+ Million	-Xmx30g	60GB	64 GB

Memory Allocation Rules

1. Keep ANALYSER_XMX ≤ 70% of container memory
2. ANALYSER_XMX scales with lines of code (LOC)
3. ANALYSER_XMX and RISK_XMX are independent settings

Configuration Examples

Server Mode — Docker Compose

version: '3.8'
services:
  app:
    image: "browserstack/code-quality:1.9.36.0"
    environment:
      - ANALYSER_XMX=-Xmx15g
      - RISK_XMX=-Xmx1024m
    deploy:
      resources:
        limits:
          memory: 30G

Server Mode — Docker Run

docker run -m 30GB -d \
  -e ANALYSER_XMX=-Xmx15360m \
  -e RISK_XMX=-Xmx1024m \
  -p 3000:3000 \
  --name BrowserStackCodeQuality \
  [other options] \
  browserstack/code-quality:1.9.36.0

CLI / Remote Analysis Mode

When running the embold-scanner CLI directly (on your machine or a CI/CD agent), set the environment variable in your shell before executing the scan:

# Set analyzer heap size for CLI analysis
export ANALYSER_XMX=-Xmx8g

# Then run the scanner as usual
./browserstack-codequality-scanner/bin/embold-scanner analyse \
  -u https://demo.embold.io \
  -t $EMBOLD_TOKEN \
  -r {YOUR_REPO_UID} \
  -c ./repository-configuration.json

In a CI/CD pipeline (e.g., GitHub Actions, Jenkins, GitLab CI), set it as a pipeline environment variable:

# GitHub Actions / GitLab CI example
env:
  ANALYSER_XMX: "-Xmx8g"

// Jenkinsfile example
environment {
    ANALYSER_XMX = '-Xmx8g'
}

Note: For CLI mode there is no container memory limit (-m). Ensure the machine or CI/CD agent has enough physical RAM to accommodate the heap size you configure.

Memory Distribution Guidelines

Analyzer Memory as % of Container Memory

Repository Size	Analyzer Memory %
< 1M LOC	35-37%
1-3M LOC	40-42%
3-5M LOC	50-55%

Memory Consumers

Primary consumers of container memory:

• Analysis Engine: 50-70% (configured via ANALYSER_XMX)
• UI Services: 15-20%
• Controller Processes: 10-15%
• Database Operations: 5-10%

Optimization Strategies

For Complex Codebases

High code duplication or mixed languages may require:

• 20-30% more memory than standard recommendations
• Additional monitoring during initial scans
• Iterative tuning based on actual performance

Performance Monitoring

# Monitor container memory usage
docker stats BrowserStackCodeQuality

# Check for memory pressure in logs
docker logs BrowserStackCodeQuality | grep -i "memory\|heap\|gc"

Troubleshooting

Out of Memory Errors

Symptoms:

• Analysis fails with heap space errors
• Container restarts during analysis
• Slow performance or timeouts

Solutions:

1. Increase ANALYSER_XMX (ensure < 70% of container memory)
2. Increase container memory limit (-m parameter)
3. Enable multi-threaded execution with appropriate memory
4. Use remote scanning for very large repositories

Memory Optimization Tips

1. Start with recommendations and adjust based on performance
2. Monitor first scan closely and tune accordingly
3. Use SSD storage to reduce memory pressure from I/O
4. Consider remote scanning for repositories > 10M LOC
5. Enable garbage collection logging for detailed analysis

Advanced Configuration

Remote Scanning for Large Repositories

For repositories exceeding memory capacity:

# Enable remote scan mode
-e EMB_REMOTE_SCAN=true
-e EMB_REMOTE_SCAN_STRICT=true  # For C++ accuracy

Monitoring and Alerts

# Set memory usage alerts
docker run --memory=30g --oom-kill-disable=false \
  [other options]

Environment Variables Reference

Variable	Where It Applies	Purpose	Example
ANALYSER_XMX	Server (Docker) & CLI	Analysis engine heap size	-Xmx15g
RISK_XMX	Server (Docker)	Risk calculation heap size	-Xmx1024m
EMB_ANALYSER_THREADS	Server (Docker) & CLI	Analysis thread count	8
EMB_PARSER_THREADS	Server (Docker) & CLI	Java parser thread count	4
EMB_REMOTE_SCAN	Server (Docker)	Enable remote scanning	true

---

Related Topics

• Multi-Threaded Execution - How to tune thread count for both Server and CLI modes
• BrowserStack Code Quality CLI - Running remote analysis from the CLI
• CI/CD Integration - Setting memory and thread variables in your pipeline
• Docker Deployment - Full container setup guide