EdgeX 4.0 Performance Revealed: Leaner, Smarter and Ready for the Future
By James Butcher, EdgeX Foundry Chairman and IOTech Systems Product Manager
EdgeX 4.0 (Odesa) marks a significant step forward for the EdgeX Foundry community. Alongside key feature enhancements, this release brings measurable improvements in memory efficiency and deployment flexibility. The newly published 4.0 performance report highlights the real-world impact of the latest architectural changes, offering developers and integrators a clear view of what’s changed — and why it matters.
What’s New in EdgeX 4.0?
As a reminder, EdgeX 4.0 – released in March this year – introduced several foundational updates:
- PostgreSQL replaced Redis as the default database, delivering more robust persistence capabilities.
- Core Keeper replaced Consul, providing a lightweight configuration service.
- OpenBao replaced Vault as the default secret store, aligning with the project’s focus on open-source maintainability.
- A new Time-Based Scheduling Service, for more precise control of edge operations.
- A new Python API, to broaden accessibility to a wider community.
These changes are all aligned with the community’s goal of providing and maintaining a modular, production-ready edge platform. Read my full 4.0 release blog for more details.
Key Performance Highlights
The EdgeX community gathers detailed performance metrics for each release. Given the microservices nature of the platform, users can choose exactly which services to deploy. For analysis, the community benchmarks two deployment modes of the platform:
- Full/Typical Deployment: most services, including security, are enabled.
- Minimal Deployment: streamlined for resource-constrained environments.
In a nutshell, here’s how EdgeX 4.0 compares to the previous 3.1 (Napa) release:
Full (Typical) Deployment
- Runtime memory usage reduced by 38% (from 335 MB to 206 MB).
- Docker image footprint reduced by ~10%, now ~996 MB.
- Startup time increased slightly with full security (from 38s to 40s).
Minimal Deployment
- Startup time improved (17s from 22s).
- CPU and memory usage increased slightly (71 MB from 38 MB) – attributed mainly due to PostgreSQL’s larger baseline footprint
Latency
- API response times with the API gateway rose marginally (from ~31ms to ~37ms).
- Response times without security improved slightly (from ~2.4ms to ~2ms).
- Data latency tested with the virtual device remains stable (around ~4–6ms), with no measurable overhead from security.
All stats relate to testing performed on a Dell 3200 Edge Gateway.
What Does This Mean?
The 4.0 metrics confirm that EdgeX remains a lightweight and efficient platform for edge computing. While PostgreSQL and OpenBao bring a larger footprint to minimal setups, they offer enhanced compatibility, durability, and manageability for enterprise-scale use cases.
Overall, EdgeX is evolving into a more robust, secure, and modern framework while maintaining its edge-friendly characteristics — particularly when carefully selecting services to match deployment constraints.
Commercial EdgeX Options
Note that while EdgeX is mature and remains open-source, many adopters prefer a ready-made commercial implementation. The pluggable microservice architecture allows vendors to extend or replace services to meet specific performance, protocol, or compliance needs. Check out the ecosystem pages of the EdgeX Website to find out more.
Get the Full Report
You can access the full EdgeX 4.0 performance report here. Whether you’re optimizing for a minimal footprint or scaling to full-featured industrial edge deployments, the latest data will help guide your design and deployment strategy.
