Introduction
Enterprise Architect (EA) is widely used in geographically distributed teams, often accessed over wide-area networks (WANs). While EA performs well in local environments, WAN setups can introduce significant latency, packet loss, and database throughput issues. These degrade the user experience with slow diagram rendering, long save/load times, and timeouts. Fortunately, several strategies can optimize EA’s performance across WANs.
This article explores common performance problems when using EA over WAN and provides actionable steps for improving responsiveness and stability in remote environments.
1. Understanding the Challenge
EA’s traditional client-server model involves direct communication between the EA client and a database backend. Over a WAN, every action—like expanding a package or opening a diagram—results in SQL queries being executed remotely. This chatty architecture is susceptible to latency and bandwidth constraints, especially for large models or remote users working via VPN.
2. Common WAN Performance Issues
- High latency: Every user action results in multiple round-trip queries
- Slow diagram rendering: Diagrams load progressively due to WAN lag
- Timeouts and dropped sessions: EA may hang if a query takes too long
- Inconsistent performance: Performance fluctuates with network congestion
3. Pro Cloud Server (PCS) to the Rescue
Pro Cloud Server provides a modern solution by enabling HTTP/HTTPS-based communication between EA clients and remote databases. Benefits include:
- Data caching and compression to reduce query size and frequency
- HTTPS tunneling to bypass slow VPN routes
- Support for REST APIs and integrations without direct DB access
Tip: Always configure PCS with SSL for security and stability.
4. Best Practices for Optimizing EA Over WAN
- Use Pro Cloud Server: Avoid direct DB access over WAN
- Deploy EA clients close to PCS instances: Minimize cross-continent hops
- Segment large models: Split large repositories into manageable components using model views or separate DBs
- Reduce diagram complexity: Minimize number of elements and relationships per diagram
- Disable auto-refresh features: Especially on shared diagrams or dashboards
5. Infrastructure Recommendations
- Ensure 1 Gbps internal LAN and 20+ Mbps WAN connectivity per user
- Deploy PCS on a server with SSD storage and at least 16 GB RAM
- Use PostgreSQL or SQL Server for large models with >1 million objects
6. Monitoring and Diagnostics
Monitor performance with tools like:
- Wireshark or NetFlow: Analyze latency and packet loss
- EA logs: Enable verbose logging in EA to trace delays
- PCS logs: Identify request bottlenecks or DB errors
7. Case Study: Cross-Region Modeling Success
A European bank with teams in Brussels and Singapore implemented PCS and moved from direct SQL access to HTTPS. Diagram load times dropped from 12 seconds to 2 seconds, and model corruption events were eliminated entirely. The switch also enabled Prolaborate dashboards without impacting core modeling performance.
Conclusion
Using EA across a WAN doesn't have to be a painful experience. With the right combination of infrastructure, deployment strategy, and configuration—especially leveraging Pro Cloud Server—teams can enjoy smooth, stable access to EA from anywhere in the world.
Enterprise Architect, Sparx EA, WAN Optimization, EA Performance, EA over VPN, Pro Cloud Server, EA Network Latency, EA HTTPS Tunneling, EA PCS Deployment, Remote EA Access, EA Bandwidth, EA WAN Troubleshooting
If you’d like hands-on training tailored to your team (Sparx Enterprise Architect, ArchiMate, TOGAF, BPMN, SysML, or the Archi tool), you can reach us via our contact page.