Interview: Insights into Dubbo-go Development, Cloud‑Native Microservice Governance, and Open‑Source Contributions

The interviewee explains that his commonly used name originated from a WeChat nickname adopted in 2012 and later used as a work alias at Alibaba.

With over eleven years in server‑side infrastructure R&D, he has worked on RPC, NoSQL storage, real‑time monitoring, instant messaging, and message push, and has contributed early research on RocksDB, gogo protobuf, and Pulsar.

Starting with C/C++, he self‑learned Go in 2013, contributed to Redis since 2015, refactored Muduo with C++11, built dubbogo, and created PikaPort for data proxy between Pika and Redis, which has been adopted by several large Chinese tech companies.

He now focuses on promoting the dubbogo community and engages in technical exchanges within the instant‑messaging field, leveraging seven to eight years of practical IM project experience.

Since late 2018, he has been fully dedicated to building the dubbogo open‑source project (https://github.com/apache/dubbo-go), which he sees as the cornerstone for Dubbo’s transition to the cloud‑native era and a key part of his instant‑messaging technology stack.

The project has attracted over 35 registered users across diverse industries such as e‑commerce, transportation, gaming, finance, IoT, healthcare, and video platforms, and the community now includes more than 100 contributors, 23 Apache Dubbo committers, and 5 PMC members.

Dubbogo implements a Go‑language version of Dubbo, fully compatible with the Java implementation; it recreates the Netty network library (getty) and the Hessian2 serialization protocol in Go, and maintains a self‑contained stack from the transport layer up to the service layer.

Its cloud‑native mission leverages Go’s strengths to extend Dubbo’s capabilities, including an early implementation of a Kubernetes‑based service registry, which later evolved into Dubbo’s official k8s registry solution.

The project also aims to bridge the gap between Java and Go, achieving compatibility with gRPC and REST‑compatible frameworks such as Spring Cloud Alibaba.

Current community efforts focus on building an ecosystem matrix, highlighted by the recently migrated Apache gateway project dubbo‑go‑pixiu, which aims to interconnect Dubbo, Spring Cloud, gRPC, and RocketMQ.

When implementing technical solutions, he emphasizes evaluating technical and customer value, breaking down goals, assembling reliable engineers, and tracking progress, while promptly surfacing risks to stakeholders for timely mitigation.

He discusses architectural challenges faced in previous roles, such as building lightweight real‑time monitoring, log collection, instant messaging, and KV storage systems from scratch, and how leveraging past experience and mature solutions helped deliver these systems efficiently.

In technology selection, he stresses balancing rapid implementation, system stability, and cost efficiency from a business perspective.

Regarding cloud‑native trends, he favors dubbogo for its maturity and also highlights multi‑cluster Kubernetes management projects like Gardener as promising for cross‑cloud deployments.

At the recent GIAC conference, he presented on dubbogo 3.0, covering new service registration, routing, communication protocols, capacity and resource evaluation in cloud‑native environments, adaptive throttling, adaptive load balancing, and proxy‑less service mesh, as well as internal Alibaba use cases.

He concludes by expressing hope that both dubbogo and GIAC will continue to grow, using the conference platform to raise awareness of dubbogo’s capabilities and drive the advancement of microservice technology in China.