Delivery: fall late October 2017
This fall's first EdgeX release is focused on delivering a minimum viable product (MVP) for further accelerating community contributions. It ould would not be a product ready for mission-critical use cases, but it is the hope of the EdgeX community that after with the fall Barcelona release, the shape of the architecture, APIs, etc. are such that the larger IoT development community will feel comfortable exploring and using EdgeX is in upcoming IoT projects.
Release Themes & Objectives
...
- Supported OS
- Windows 10 (latest 2016 version)
- Linux, Ubuntu 16.04 classic (but does not enforce secure boot)
- EdgeX supports and welcomes other OS providers to test and validate EdgeX works on their OS
- Example, Canonical plans to test/validate EdgeX on Ubuntu Core 16
- Supported HardewareHardware:
- Intel
- EdgeX welcomes and will support 3rd party contribution of Arm support. Work to port to Arm is nearing completion.
- Performance targets stand, but may not be achieved in this release
- EdgeX to develop performance tests to show current performance metrics
- No significant performance-specific work will be accomplished in this release (unless it is easy to achieve with no other impacts)
- EdgeX will provide early Go performance numbers to show intended path to achieve performance targets (that is slowly replacing Java-based micro services with Go or other faster/smaller programming langauge language micro services over the next few releases)
- EdgeX will use current scalability metrics on devices, collection, etc. as scale guidance
- More formal scalability concerns to be addressed in future releases
- All Micro Services will be hardenhardened; meaning
- Works properly for the intended use case
- It may not be 100% complete implementations for all use cases or parts of a protocol for example, but it provides enough implementation to sustain the demo use cases for Barcelona and could support extension to the full needs or protocol in the future
- Handles errors and exceptions gracefully
- Contains proper unit and integration tests
- Follows good coding standards, and is well documented
- Following some prescribed standard (like Oracle, Google or Twitter guides for Java)
- Performs within the target measures established for Barcelona
- Code base is not exploitable
- API set is solid
...