This article examines the strengths and trade‑offs of REST, GraphQL, and tRPC for API design in 2025, helping teams decide which architecture aligns with their expertise, project scope, and performance requirements.
This article walks frontend developers through the fundamentals of the TRPC framework, explaining why understanding RPC protocols, the layered architecture, plugin system, multi‑process handling, and Protobuf serialization is essential for building high‑performance, scalable services and improving debugging efficiency.
Tencent Docs tackles the trade‑offs of monolithic and microservice architectures by implementing a flexible, automated system that merges selected microservices into a few monoliths for private‑cloud scenarios, reducing runtime costs, deployment overhead, image size, memory usage and CPU consumption while preserving performance and scalability.
This article explains how Tencent Docs tackled the high runtime, deployment, and image‑distribution costs of a large microservice fleet by building a flexible architecture that can merge selected services into a few monolith binaries for private deployments, while retaining microservices for public scenarios, and shares the tooling, configuration, challenges, and performance gains achieved.
This article explores how React is evolving into a full‑stack framework by adding server components, server actions, and leveraging TypeScript and tRPC, while reflecting on the shift from client‑side rendering to a unified front‑end/back‑end development model.
The article explores how React has evolved with server components, server actions, and TypeScript to bridge front‑end and back‑end development, discusses the shift from CSR to SSR, and examines the impact of tRPC and generated type interfaces on modern web engineering.
This article recounts the technical and organizational challenges of building Tencent WeSee's 2020 Spring video red‑packet system, detailing high‑traffic spikes, consistency and security demands, TRPC adoption risks, performance bottlenecks, and the 33 robustness principles that guided the subsequent architecture overhaul.
By introducing the unified tRPC scaffolding tool trpcx and embedding OpenTelemetry‑generated observability configurations, the Tencent Docs backend team streamlined service creation, standardized directory structures, migrated metrics and logs to ClickHouse for cost‑effective performance, and established best‑practice workflows that dramatically improve development speed and fault‑diagnosis efficiency.
The article details how a Node‑based business gateway for QQ Channel services replaced legacy RPC with a lightweight tRPC solution, achieving ten‑fold latency reduction, higher QPS, improved security and availability, and added observability through logging, metrics, and a WebSocket push layer.
This article shares a decade‑long journey of designing and scaling the QQ front‑end unified access layer, covering business background, overall architecture, solution comparisons, core challenges, observability, and performance optimizations while reflecting on practical lessons for large‑scale front‑end systems.
The veteran front‑end engineer chronicles a decade of building QQ’s large‑scale products, detailing how the new Frontend Unified Access Layer replaced fragmented SDKs with a high‑performance, scalable, secure gateway built on an internal http2rpc framework, while tackling legacy protocol coexistence, observability, alert fatigue, and targeted performance optimizations.
This article details how Tencent Docs implements a flexible architecture that can switch between monolithic and microservice deployments, describing the motivations, such as framework diversity, configuration conflicts, global variable management, hidden bugs, the custom 'monolith' tool with its configuration format, and the resulting performance and resource savings.
Tencent Docs built a configurable “monolith” tool that merges selected tRPC‑Go microservices into a few monolithic modules for private‑cloud deployments, overcoming framework, configuration, global‑state, and bug‑resolution challenges, and achieving up to 75 % image‑size reduction and 96 % memory savings while preserving microservice advantages where needed.
This article explains how the team transformed a Go‑based microservice recommendation system into a single‑pod monolithic application using tRPC‑Go, detailing performance bottlenecks, code‑level mock‑proxy techniques, deployment adjustments, and the resulting dramatic reduction in CPU consumption.
This article explains how a large‑scale recommendation system built with tRPC‑Go microservices was refactored into a single‑process monolith to cut network overhead, reduce CPU usage by over 60%, and retain the benefits of microservice development while minimizing code changes.
The article shows how to monolithize selected tRPC‑Go microservices by defining protobuf‑generated Go interfaces and swapping RPC proxies for in‑process implementations via a proxy API, cutting CPU usage by 61% while keeping microservice flexibility and offering best‑practice guidelines for Go service design.
tRPC is an open‑source, plugin‑based RPC framework that separates core communication, service‑governance, and invocation layers, supports extensible codecs, naming, config, metrics, logging and tracing plugins, offers an Admin API, and demonstrates high throughput and low tail latency in cloud‑native deployments.
Tencent engineers completely rewrote a ten‑year‑old query optimizer, shrinking the codebase by 80% and replacing its monolithic thread‑pool with a tRPC‑Fiber DAG scheduler, which cut latency by 28%, reduced startup time to five minutes, saved 40 GB memory, boosted throughput 12%, fixed numerous stability bugs, raised test coverage above 60%, and accelerated new‑feature lead time to under a day.
