// the take
A Spring Cloud microservices reference platform targeting Chinese enterprise developers — covers auth (Spring Authorization Server + OAuth2/JWT), multi-tenancy, RBAC, service mesh via Nacos/Sentinel, and a full observability stack (ELK, SkyWalking, Grafana). It's not a library you pull in; it's a runnable template you fork and extend. Aimed at teams that want a pre-wired architecture they can drop business logic into.
The auth layer is the real value here — Spring Authorization Server with custom grant types (mobile, OpenId, password+code) on top of stateless JWT is non-trivial to wire correctly and this does it properly. The commons starters (zlt-auth-client-spring-boot-starter, zlt-redis-spring-boot-starter, etc.) are genuinely reusable isolation points rather than dumping everything in one module. The observability setup is complete out of the box — SkyWalking for APM, ELK for logs, Grafana/Prometheus for metrics, plus a slow-SQL monitor — most similar projects ship one of these, not all four. Multi-tenant app isolation is baked into the data model from the start, which is far easier than retrofitting it.
Spring Cloud 2022.x still pins Elasticsearch 7.x when 8.x has been out for years — the ES client API changed significantly and that's a real migration cliff. Documentation is almost entirely in Chinese with no English equivalent beyond a stub README, which walls off non-Chinese contributors and makes it hard to evaluate architectural decisions from the outside. The repo history shows the last meaningful push was April 2025 on the master branch, but it's unclear whether it's maintained or effectively archived — issues and PRs aren't visibly triaged. Zuul is still referenced in the topics even though the gateway is now Spring Cloud Gateway — the leftover reference suggests the project's own metadata hasn't been cleaned up, which doesn't inspire confidence in internal consistency.
We read the README, the directory tree and a key source file in June 2026. Nobody pays to be listed. How we judge →