Gaode Maps' Large‑Scale Serverless Deployment for the 2020 Golden Week Travel Festival

Serverless has transitioned from a speculative technology to large‑scale production. During the 2020 “Golden Week” travel festival, Gaode Maps (AutoNavi) applied Serverless to its core autonomous travel services, handling over 100 million active users and reaching nearly two million function invocations per minute.

The autonomous travel business includes the main map page, route‑planning page, and navigation‑end page. Gaode stored UI templates in the cloud and used Node FaaS (Alibaba Cloud Function Compute) to fetch templates, aggregate data, process logic, and generate schema. The client renders the UI directly from the schema, enabling rapid iteration and high flexibility.

Performance metrics showed a service success rate above 99.99%, more than 1 million triggers per minute (QPS > 20 k), and average response times under 60 ms across all scenarios, exceeding expectations during peak traffic.

Business value – Efficiency : Replacing traditional BFF layers with Serverless For Front‑end (SFF) eliminates code bloat, reduces operational overhead, lowers costs, and improves resource utilization. Gaode’s end‑to‑end lifecycle (development, testing, gray release, production) became 40 % faster, significantly boosting developer productivity.

Business value – Elasticity & Cost : Function Compute automatically scales with traffic, avoiding over‑provisioning during low‑traffic periods and reducing infrastructure waste. Millisecond‑level cold‑start times enable rapid scaling up and down, making elastic resource usage accessible to large enterprises.

Business value – Observability : Integration with Alibaba Cloud’s logging, monitoring, tracing, and workflow services provides full‑stack observability with a single configuration, simplifying performance diagnosis and reducing learning costs.

Gaode’s large‑scale Serverless rollout demonstrates the practical benefits of cloud‑native architectures: faster business iteration, enhanced flexibility, and a solid foundation for future cloud migration across various industry scenarios.