Red Hat Migration Toolkit for Applications (MTA) - AI-Optimized Technical Reference

Core Function

Java application migration analysis tool that identifies containerization and modernization blockers through static code analysis. Formerly called Windup.

Critical Deployment Intelligence

Working Installation Options

• CLI Tool: Only reliable option for enterprise codebases
• IDE Extensions: Unstable but provide inline warnings when functional
• OpenShift Operator: Requires expensive OpenShift subscription, web UI crashes on real applications

System Requirements (Real-World)

• Memory: 16GB minimum (official 8GB causes heap overflow)
• Analysis Time: 15 minutes for toy apps, 6+ hours for enterprise codebases
• JVM Settings: WINDUP_OPTS="-Xmx24g -XX:+UseG1GC" required for large applications

Migration Analysis Capabilities

Supported Migration Paths

Source	Target	Reliability	Common Issues
WebLogic → JBoss EAP	High pain migration	Finds major compatibility issues	Buried critical issues in 500-page reports
Ancient JBoss → Modern EAP	Medium complexity	Catches deprecated EJB patterns	Rule updates break custom configurations
Oracle JDK → OpenJDK	Low-medium effort	Identifies vendor-specific APIs	Misses Maven parent POM issues
Camel 2 → 3/4	Framework-dependent	Maps breaking changes	False positives on standard patterns
General containerization	Variable success	Finds hardcoded paths/configs	Overwhelming false positive rate

Analysis Engine Performance

• Rule Base: 1000+ predefined rules, 90% generate false positives
• Scanning Modes: Source code, JAR files, EAR archives
• Memory Loading: Loads entire Maven repository into memory simultaneously
• Network Dependencies: Fails behind corporate firewalls without proxy configuration

Critical Failure Scenarios

Memory and Performance Issues

# Common failure
Exception in thread "main" java.lang.OutOfMemoryError: Java heap space

# Required workaround
export WINDUP_OPTS="-Xmx24g -XX:+UseG1GC"

Windows Environment Blockers

• Path length limit: 260 characters breaks Maven projects with deep hierarchies
• Error: The filename or extension is too long
• Solution: Use WSL2 or Docker Desktop for longer path support

Network Configuration Failures

• Corporate firewalls block dependency resolution
• Error: Could not resolve dependencies for project
• Required: Maven proxy settings and firewall exceptions for Red Hat repositories

Report Analysis Intelligence

Issue Categories by Priority

Category	Action Required	Reliability
MANDATORY	Fix immediately	70% accurate
CONNECTION POOL	Database migration blocker	85% accurate
HARDCODED	Container deployment blocker	60% accurate
INFORMATION	Ignore completely	5% useful
OPTIONAL	Red Hat product sales pitch	0% useful
POTENTIAL	False positive noise	10% accurate

Effort Estimation Reality

• MTA Estimates: Fictional, based on hello-world assumptions
• Real Multiplier: 3x minimum for any estimate
• Story Points: Ignore completely, assume perfect documentation doesn't exist

Tool Comparison Matrix

Tool	Java Analysis	Binary Support	Container Assessment	Custom Rules	Pricing Model
Red Hat MTA	Deep static analysis	✅ Decompilation	OpenShift-specific	✅ XML-based	OpenShift subscription
AWS App Discovery	❌ Infrastructure only	❌ Not supported	❌ Not included	❌ Not supported	Pay-per-discovery
Azure Migrate	🔶 Limited	❌ Not supported	🔶 Basic	❌ Not supported	Azure subscription
IBM Transformation Advisor	WebSphere focus	✅ WebSphere binaries	🔶 Limited	🔶 Limited	WebSphere license

Operational Workarounds

CLI Usage (Reliable Method)

# Wrapper script required for enterprise use
windup-cli --input /path/to/app --output /path/to/results --verbose
# Always use --verbose or get mystery failures

Custom Rules Development

• Format: Baroque XML syntax similar to XSLT
• Learning Curve: 2-3 weeks to become productive
• Breaking Changes: Rule updates break custom rules without warning
• Strategy: Copy existing rules from /usr/share/windup/rules/ and modify

Report Filtering Strategy

1. Search for "MANDATORY" issues first
2. Focus on CONNECTION POOL and HARDCODED categories
3. Ignore everything marked "OPTIONAL" or "INFORMATION"
4. Expect 90% false positive rate on binary analysis

Resource Investment Requirements

Learning Curve

• Initial Proficiency: 2-3 weeks to understand quirks and limitations
• Advanced Usage: Additional 2-4 weeks for custom rule development
• Team Training: Multiply by team size, expect knowledge transfer friction

Infrastructure Costs

• OpenShift Subscription: Thousands per month for enterprise scale
• Hardware Requirements: 16-32GB RAM servers for analysis
• Support Costs: Red Hat support required for enterprise deployment issues

Critical Warnings

What Official Documentation Doesn't Tell You

• Web console designed for demo applications, not enterprise codebases
• Memory requirements severely underestimated for real-world usage
• Rule engine updates break existing custom configurations
• IDE plugins crash frequently, save work often
• Windows path limitations break most enterprise Maven projects

Breaking Points and Limitations

• UI Timeout: Web interface fails on applications >100MB
• Memory Ceiling: Heap overflow on codebases >500K lines without tuning
• Rule Reliability: High false positive rate makes manual filtering mandatory
• Framework Support: Biased toward Red Hat solutions, poor Spring Boot analysis

Migration Project Reality Check

• Projects take 3x longer than estimated
• "Simple" applications have undocumented dependencies
• ClassPath issues emerge during runtime, not analysis
• Team expertise requirements exceed tool documentation assumptions

Decision Criteria

Use MTA When

• Migrating from WebLogic/WebSphere to JBoss EAP
• Large codebase needs systematic dependency analysis
• Team has 2-3 weeks to learn tool limitations
• Budget includes OpenShift subscription or CLI-only usage acceptable

Avoid MTA When

• Modern Spring Boot applications (high false positive rate)
• Small codebases (<50K lines, manual review more efficient)
• Windows-only development environment
• No Red Hat ecosystem investment planned

Alternative Assessment

• Binary-only analysis: Limited value, high false positive rate
• IDE integration: Useful for new development, unreliable for legacy cleanup
• Custom rule development: High investment, frequent breaking changes
• Automated fixes: Essentially non-existent, manual remediation required

Success Patterns

Effective Usage Strategy

1. Start with simplest application to learn tool patterns
2. Use CLI exclusively, avoid web interface
3. Develop custom rules for organization-specific anti-patterns
4. Filter reports aggressively, focus on blocking issues only
5. Plan 3x time multiplier for all effort estimates

Integration Approaches

• CI/CD: Jenkins pipeline integration works better than web console
• Automation: Wrapper scripts required for enterprise filtering
• Team Workflow: Designate MTA expert to filter noise for development teams
• Reporting: Generate CSV outputs for management tracking vs HTML dumps