Checklist: Preparing Your AEM 6.5 Environment for an Upgrade

A comprehensive checklist to ensure your AEM 6.5 environment is ready for upgrade, covering discovery, testing, deployment, and post-upgrade validation.

Upgrading Adobe Experience Manager is a complex undertaking that requires careful preparation. This checklist will help you ensure nothing is overlooked as you plan and execute your AEM 6.5 upgrade.

Phase 1: Discovery & Assessment

Environment Inventory

Document all AEM instances (author, publish, dispatcher)
List all integrations (CRM, analytics, DAM, etc.)
Catalog third-party packages and versions
Identify all custom OSGi bundles and components
Map out content structure and templates
Document current system architecture

Custom Code Audit

Review all custom components for deprecated APIs
Check servlets and services for compatibility
Audit JavaScript and client libraries
Review custom workflows and launchers
Identify unused or redundant custom code
Check for hardcoded paths or configurations

Performance Baseline

Document current page load times
Record repository size and growth rate
Note peak traffic patterns and response times
Benchmark query performance
Document current resource utilization (CPU, memory, disk)

Team Preparation

Identify key stakeholders and decision-makers
Assign upgrade project roles and responsibilities
Schedule regular status meetings
Plan for knowledge transfer sessions
Identify training needs for new features

Phase 2: Planning

Technical Planning

Choose target AEM version (newer 6.5 LTS or AEM as a Cloud Service)
Plan upgrade path (direct vs. phased)
Design new architecture if making improvements
Plan for dispatcher and CDN updates
Determine database upgrade strategy
Plan Java version upgrade if needed

Testing Strategy

Define test scenarios covering all major features
Identify critical user journeys to validate
Plan for performance testing
Schedule user acceptance testing (UAT)
Prepare test data and content
Set up test environments matching production

Risk Mitigation

Document rollback procedures
Plan for blue-green or canary deployment
Identify potential breaking changes
Create backup and restore procedures
Plan communication strategy for stakeholders
Prepare contingency plans for common issues

Resource Planning

Secure budget for upgrade costs
Allocate internal team time
Engage external consultants if needed
Schedule downtime windows
Plan for post-upgrade support coverage

Phase 3: Development & Testing

Environment Setup

Clone production to lower environments
Set up development environment with new AEM version
Configure staging environment for testing
Implement monitoring and logging
Set up automated deployment pipelines

Code Migration

Upgrade custom components to new APIs
Update deprecated code patterns
Refactor problematic code identified in audit
Update client libraries and dependencies
Migrate configurations to new format if needed
Update documentation for code changes

Integration Testing

Test all third-party integrations
Validate authentication and SSO
Check data synchronization processes
Test API endpoints and web services
Verify CDN and dispatcher configuration
Validate analytics and tag management

Content Migration

Test content migration scripts
Validate all content renders correctly
Check for broken references
Verify asset renditions
Test taxonomy and metadata
Validate multilingual content if applicable

Performance Testing

Run load tests matching production traffic
Test under peak load conditions
Validate caching behavior
Check query performance
Monitor resource utilization
Identify and fix performance bottlenecks

User Acceptance Testing

Execute all planned test scenarios
Validate critical business processes
Test across different browsers and devices
Verify accessibility standards
Check for visual regressions
Document any issues found

Phase 4: Deployment

Pre-Deployment

Final backup of production environment
Freeze content changes if possible
Communicate maintenance window to users
Prepare deployment runbook
Assign on-call support team
Set up monitoring and alerting

Deployment Steps

Execute deployment according to runbook
Validate each step before proceeding
Monitor system health during deployment
Perform smoke tests after each major step
Document any deviations from plan
Keep stakeholders updated on progress

Post-Deployment Validation

Run automated health checks
Validate critical functionality
Check all integrations are working
Verify content is intact
Test user authentication
Validate scheduled jobs and workflows
Check system performance metrics
Review logs for errors or warnings

Phase 5: Post-Upgrade

Monitoring

Monitor error logs closely for 48-72 hours
Track performance metrics vs. baseline
Watch for user-reported issues
Monitor system resource utilization
Check for memory leaks or performance degradation
Validate backup processes are working

Documentation

Document the upgrade process and lessons learned
Update architectural diagrams
Record configuration changes
Update runbooks and procedures
Document known issues and workarounds
Create knowledge base articles for team

Knowledge Transfer

Train team on any new features used
Share post-upgrade findings with stakeholders
Conduct retrospective meeting
Update onboarding materials for new team members
Document any new monitoring or alerting

Optimization

Review performance and identify improvements
Optimize configurations based on monitoring
Clean up unused code or content
Implement any deferred optimizations
Schedule regular health checks

Common Pitfalls to Avoid

Don’t Skip Testing

Testing in lower environments is not optional. Many issues only appear under certain conditions or with specific content.

Don’t Underestimate Timing

Most upgrades take longer than estimated. Build in buffer time for unexpected issues.

Don’t Forget About Dispatcher

The dispatcher configuration often needs updates. Test it thoroughly.

Don’t Ignore Deprecated Features

Just because deprecated code still works doesn’t mean it will continue to work in future versions.

Don’t Deploy Without Rollback Plan

If something goes wrong, you need a quick way back to the working state.

How Long Does Each Phase Take?

Based on a typical enterprise AEM implementation:

Discovery & Assessment: 2-3 weeks
Planning: 2-4 weeks
Development & Testing: 4-8 weeks
Deployment: 1-3 days
Post-Upgrade Stabilization: 1-2 weeks

Total: Typically 12-18 weeks from start to finish for a well-planned upgrade.

Need Help?

Don’t want to tackle this alone? We’ve executed dozens of AEM upgrades and can help at any stage:

Health Check: Assess your readiness and create a plan
Execution: Handle the entire upgrade process
Advisory: Guide your team through the upgrade
Support: Provide post-upgrade support and optimization

Download this checklist as a PDF to share with your team. Get the PDF (Coming soon)