Gaode’s Cloud Migration: Operational Transformation and Best Practices on Alibaba Cloud

Cloud computing brings undeniable transformation; as a new IT delivery model it saves costs, accelerates business‑IT integration, enhances innovation, management, and reliability, becoming a mainstream path for enterprise development.

In an interview, Gaode’s senior operations expert Zeng Wei discusses the cloud journey, benefits, architecture, and the shift in operational value that raised service availability to 99.99%, a five‑fold increase, while cutting operations staff from 45 to 16.

Gaode’s IT‑Informationization Roadmap

Founded in 2002, Gaode evolved from simple car‑navigation products to a comprehensive internet platform offering location, search, routing, navigation, O2O services, vehicle‑networking, and real‑time traffic data.

The IT evolution progressed through three stages:

Chaotic phase without an operations team (2002‑2007): Servers were managed by the IT department, with limited online map coverage.

Establishment of a professional operations team (2007‑2014): IT handled office infrastructure while operations managed IDC, servers, and business services, still largely manual and inefficient.

Cloud‑based automation (July 2014‑present): Migration to the cloud enabled automated operations and improved fault handling and performance.

Gaode’s Data Collection, Processing, and Integration Process

The data pipeline has evolved through four stages:

Initial stage: Manual collection via cameras, GPS, video discs, with three‑month update cycles.

System stage: Introduction of collection software and networked upload, moving away from physical media.

Platform stage: High‑precision collection vehicles, crowdsourced data, and platform‑based workflow with full monitoring.

Automation stage: Future crowdsourced, automated back‑haul, and AI‑driven image recognition and big‑data analysis for fully automated production.

Gaode now offers map and navigation apps, PC maps, and open APIs for developers.

Best Operational Practices on Alibaba Cloud

Operational pain points before cloud migration: Distributed small data centers caused power, cooling, and network failures, requiring over 40 staff; diverse responsibilities across IDC, OS, databases (MySQL, Redis, Oracle, etc.), middleware (Nginx, Apache), and tools (Jenkins, Puppet) created high pressure and occasional outages during traffic spikes.

Alibaba Cloud solutions: Unified infrastructure services (ECS, RDS, OSS, Redis, MaxCompute, etc.) eliminated the need to maintain underlying components. Elastic scaling handled traffic surges, allowing operations to focus on architecture, performance, and incident response.

Initial cloud adoption challenges included unfamiliarity with cloud components and latency in log transmission between Gaode IDC and Alibaba Cloud, which were resolved through Alibaba’s technical support and training.

During MySQL migration to RDS, Gaode used Alibaba’s Data Transmission Service (DTS) for seamless cross‑network migration and leveraged the Data Replication Center (DRC) for real‑time heterogeneous database synchronization.

Post‑migration, service availability rose to 99.99%, a five‑fold improvement over the previous seven small data centers. Deployment time shrank from months to minutes; elastic scaling now completes in ~20 minutes, and new applications launch within 10‑20 minutes.

All core business systems now run on Alibaba Cloud, removing the need for on‑premise IDC, network, and server maintenance. Cloud services cover databases, caches, load balancers, and deployment tools, while multi‑region cloud resources provide disaster recovery. Security services (DDoS protection, traffic cleaning, anti‑virus, fraud detection, code scanning) further harden the platform.

IaaS: Multi‑region data centers, networking, and inspection.

PaaS: Elastic Compute (ECS), RDS, OSS, load balancing, CDN, MaxCompute.

SaaS: Gaode’s own systems delivered as services.

The shift from maintaining physical infrastructure to consuming cloud services transformed operational focus to architecture, fault handling, and performance optimization, reducing staff from 45 to 16 and increasing job satisfaction.

Beyond navigation, Gaode now exploits Alibaba Cloud’s “Data Plus” platform—MaxCompute, Stream Computing—to perform real‑time traffic analysis, congestion avoidance, hotspot detection, and predictive modeling using location data, supporting intelligent transportation, vehicle‑networking, and future autonomous driving initiatives.