cvetomir-todorov / cecochat

Chat engine designed for millions of users. System design based on microservice architecture. Implementation based on cloud agnostic technologies. Configurable, observable, containerized and runnable, deployable local environment. Kafka, YugabyteDB, Cassandra, MinIO, HTTP, gRPC, Redis, .NET 8, ASP.NET, OpenTelemetry, Docker, Kubernetes.

Chat engine designed for millions of users

• Messaging in real-time
  • Send and receive messages
  • Send files - images, text, PDF
  • React/unreact with emojis
  • Multiple clients for the same user
• Chats
  • Notifications when a message has been processed
  • Indication for new messages
  • Review history at a random point in time
• User
  • Register, authenticate
  • Change password, edit trivial profile data
  • Store small user files - images, text, PDF
• Other users
  • Search other users by name
  • Connect with other users - invite/accept/cancel/remove
  • Profiles with static full/public-only data

Check out what next needs to be implemented. I appreciate all comments so feel free to use the Discussions tab.

Code

Technology

• System design based on microservice architecture
• Implementation based on cloud agnostic technologies
• Configurable, observable, containerized and runnable, deployable local environment
• Main technologies
  • Integration - Kafka, gRPC, WebSocket, HTTP
  • Data storage - Yugabyte, Cassandra, MinIO, Redis
  • Operations - OpenTelemetry, Docker, Kubernetes
  • Services - .NET, ASP.NET, SignalR, EF Core
  • Libraries - Autofac, Serilog, FluentValidation, AutoMapper, Polly, IdGen

Design

Documentation

• Intro
  • Capabilities
  • Overall design
  • Technologies
• Research
  • Concurrent connections limit
  • Calculations
  • Messaging traffic
  • Message IDs
  • Reliable messaging and consistency
• Design
  • Messaging
  • Chats
  • User - profiles, connections, files
  • Clients
  • Configuration - static and dynamic
  • Observability
  • Deployment
• Development
  • Local run prerequisites
  • Local run in Docker
  • Local run in Minikube
  • Development process
• Load test using 2 machines
• What next

Ban

Any use of the project for training large language models, generative AI, or any other similar tools in the past/present/future is done without our permission.

License

• Server Side Public License (SSPL)
• FAQ about SSPL
• Comparison of AGPL to SSPL