Introduction
Sparx Enterprise Architect (EA) is a robust modeling tool, but like any enterprise repository system, it can suffer from performance degradation, orphaned elements, or data integrity issues over time—especially in collaborative environments. EA’s built-in Project Integrity Check is a powerful utility that scans the repository for inconsistencies and offers the option to repair them. Yet many users overlook this tool or run it improperly.
This article explores best practices for using EA's Project Integrity Check to maintain a clean, healthy modeling environment and ensure consistent behavior across teams and integrations.
1. What Does the Integrity Check Do?
The Integrity Check tool performs a scan of the EA repository and identifies:
- Orphaned elements (objects not associated with any diagram or package)
- Dangling connectors (references to non-existent elements)
- Corrupt stereotype or tagged value definitions
- Missing or incorrect references in system tables
The tool then allows you to select and apply fixes either individually or in bulk.
2. When and Why to Run It
Regular use of the Integrity Check is critical in collaborative or large-scale projects. Recommended triggers include:
- Before exporting models to XMI for team exchange
- After model merges or Pro Cloud Server integration
- Following a database migration or upgrade
- After restoring from backup
- Prior to baselining or auditing model content
3. How to Run the Integrity Check
- Open your EA project.
- Go to Configure > Integrity > Project Integrity
- Select which tests to perform (or use Check All)
- Click Run to begin scanning
- Review the issues and choose to Fix Selected or Fix All
Tip: Always back up your .eap or DB repository before fixing issues.
4. Common Issues Detected
- Orphaned Elements: Common when packages are deleted without cleanup
- Dangling Connectors: Often result from manual database manipulation
- Unused Attributes/Operations: Especially after reverse engineering
- Corrupted Stereotypes: Resulting from imported MDG technologies or profile mismatches
5. Interpreting Results Carefully
Some "errors" are expected and context-dependent. For example, orphaned elements might be deliberate placeholders. Never fix without understanding the context, especially in shared repositories.
6. Integrity Checks in SQL-Based Repositories
For EA models hosted in SQL Server, PostgreSQL, or Oracle, always ensure:
- Read/write access to the DB during integrity operations
- No conflicting access or locks during cleanup
- Integrity check is run outside of peak hours to avoid performance issues
7. Pro Tips for Ongoing Health
- Automate monthly integrity scans using EA scripting API
- Use Prolaborate dashboards to flag potential model inconsistencies
- Encourage modelers to clean up as they go—adopt modeling standards
Conclusion
The Project Integrity Check is EA’s unsung hero for repository hygiene. With regular use, clear interpretation of results, and informed application of fixes, you can significantly improve model quality, stability, and trustworthiness. Clean models aren’t just easier to work with—they’re more resilient to integration, export, and governance processes.
Keywords
Enterprise Architect, EA Integrity Check, Sparx EA Best Practices, EA Repository Health, EA Data Cleanup, EA Orphaned Elements, EA Dangling Connectors, EA Corruption Repair, EA Maintenance, Prolaborate, EA Model Governance, EA Project Tips