Our team is debating between using traditional traceability matrices versus LiveDocs for managing requirements traceability across a complex aerospace system with 8000+ requirements spanning multiple subsystems. We currently use matrices but they’re becoming unwieldy - slow to load, difficult to navigate, and the auto-refresh behavior causes performance issues during large updates. LiveDocs seems more flexible for presenting traceability relationships, but we’re concerned about scalability and whether it can handle our volume. What are the real-world performance metrics people are seeing? Has anyone successfully transitioned from matrices to LiveDocs at this scale, and what about the training needs for the team?
The hybrid approach makes sense for our compliance needs. How do you handle the auto-refresh issue in LiveDocs? Our concern is that during active development, constant refreshing of large documents will impact productivity. Can you disable auto-refresh selectively or configure different refresh policies for different document types?
Auto-refresh is configurable per LiveDoc. Set autoRefresh="false" in the document properties and add manual refresh buttons where needed. For development documents, we use a 5-minute refresh interval. For baseline documents, auto-refresh is disabled entirely. You can also configure refresh triggers based on specific work item changes rather than any change in the query scope. This gives you fine-grained control over when documents update, balancing data freshness with performance.
Performance metrics from our deployments: Traceability matrices scale poorly beyond 5000 items - we see exponential load time increases. LiveDocs with proper indexing and query optimization handle 15000+ items with linear performance degradation. Critical factors: use filtered queries rather than showing all links, implement lazy loading for nested sections, and configure caching appropriately. For aerospace compliance, ensure your LiveDocs templates include all required traceability attributes that auditors expect to see in matrices.
We made this exact transition last year with about 6500 requirements. LiveDocs performance is significantly better for viewing and navigation, but the initial setup requires careful planning. The key is structuring your LiveDocs with proper filtering and segmentation rather than trying to show everything in one document. We created separate LiveDocs for each subsystem with cross-references, which improved load times from 45 seconds (matrix) to 8-12 seconds (LiveDocs). Auto-refresh is configurable - we disabled it for large documents and use manual refresh instead.
Training needs are substantial but manageable. We developed a two-phase approach: basic training (4 hours) covering LiveDoc navigation and viewing, followed by advanced training (8 hours) for document authors covering query syntax, template customization, and performance optimization. The key insight is that most team members only need viewing skills - only 10-15% need authoring skills. Create template LiveDocs that others can clone and adapt rather than building from scratch. Document common query patterns and performance best practices in a team wiki.