Cheops - Can a "service mesh" be the right solution for the Edge?
summary
Prototype Cheops claims to make existing microservice apps edge‑compliant without changing their code by decorating APIs with a small DSL (scope + collaboration operators) and a service‑mesh layer. Implementation uses HAProxy intercept to capture native OpenStack/Kubernetes API calls, a core+glue design that talks peer‑to‑peer over RabbitMQ + AMQP, and stores metadata in ArangoDB. Sharing was demoed in 2018; replication and cross‑site POCs run on Kubernetes with eventual consistency for replicas and basic network‑partition handling.
AI-written summary · about
chapters (15)
- 0:00Start
- 3:31Existing Cloud Apps Are Going To The Edge
- 4:48Cloud Application Example: Openstock
- 8:55Same Cloud Application Instantiated Everywhere
- 10:13Focus On Collaboration (3 Types)
- 11:02My Cloud Application with Sharing
- 11:47My Cloud Application With Replication
- 13:20My Cloud Application with Cross
- 15:23Collaborations At The Devops Level
- 16:20Collaborations and Scopelang
- 22:27Behaviors Patterns
- 27:00A New Architecture
- 27:45Our Framework
- 28:52Cheops Framework
- 31:32Takeaway And Future Work
related talks
38mExtension and expansion of Service Mesh in edge computing scenarios
47mEclipse Che and Development of Microservices Apps
36mThe Service Mesh: Past, Present, and Future [B]
11mWelcome & Opening Remarks
40mComparison of Current Service Mesh Architectures
01h 03mJust what is a "service mesh", and if I get one will it make everything OK?
43mService Mesh and the future of networking
46mWhat the hell is service mesh anyway !?
