Skip to content

Microservices_Architecture_Breakdown.md

Latest commit

 

History

History
78 lines (59 loc) · 3.7 KB

Microservices_Architecture_Breakdown.md

File metadata and controls

78 lines (59 loc) · 3.7 KB

📦 Microservices Architecture Breakdown – DeathStarBench Social Network

The DeathStarBench Social Network benchmark is composed of numerous microservices, each fulfilling a specific function within the social network platform. Below is a categorized breakdown:

🧩 Core Microservices

Service Name Description
UserService Manages user registration, authentication, and profiles.
UserTimelineService Manages the user's personal timeline (posts from themselves).
HomeTimelineService Aggregates posts from followed users into the user's home feed.
ComposePostService Coordinates the creation of a new post across services.
PostStorageService Stores posts and metadata persistently.
MediaService Handles image/video uploads and storage.
TextService Processes and formats text content.
UserMentionService Identifies and links @mentions in posts.
UrlShortenService Shortens URLs included in post content.
UniqueIdService Generates unique IDs for posts, users, etc.
SocialGraphService Manages relationships like follows and followers.

💾 Supporting Services

Category Service Name(s) Role
Databases MongoDB instances Persistent storage for user, post, media, social graph data.
Caching Redis, Memcached Speed up access to frequently requested data.
Frontend media-frontend, nginx-thrift Web UI and HTTP request routing.
Observability Jaeger Distributed tracing for analyzing inter-service communication.

🔁 Typical Request Flow: Compose a Post

  1. User hits frontend (nginx-thrift) with a Compose request.
  2. ComposePostService:
  • Calls TextService to process the text.
  • Calls MediaService if an image/video is attached.
  • Calls UrlShortenService to shorten any included links.
  • Calls UserMentionService to parse and link any @mentions.
  • Calls UniqueIdService to assign a new post ID.
  • Stores the final post in PostStorageService.
  • Updates UserTimelineService and HomeTimelineService.

📦 Technologies Used per

Layer Services Technologies
API Gateway nginx-thrift OpenResty (Nginx + Lua).
Business Logic ~12 core services Go, C++, Java, Python.
Database MongoDB, Redis, Memcached Storage and Caching.
Observability Jaeger Distributed Tracing.
Load Gen wrk2 LuaJIT-based HTTP load testing tool.

🔁 Service Interactions Overview

  • ComposePostService orchestrates the workflow for new posts.
  • Posts are stored via PostStorageService.
  • Timelines are updated via UserTimelineService and HomeTimelineService.
  • All services interact with their respective Redis/Mongo/Memcached instances.
  • Jaeger is used to trace these service calls and monitor latency.

🌐 Access Points

  • Nginx Frontend: http://localhost:8080
  • Media Frontend: http://localhost:8081
  • Jaeger UI: http://localhost:16686

This architecture supports distributed, scalable, and modular deployment of a social networking platform, making it ideal for microservices benchmarking and performance analysis.