How I Grounded My First Year at Alibaba: Building OceanBase Cloud and Real‑Time Monitoring

On his first day, which happened to be the Lantern Festival, the author arrived in Beijing as the sole member of a small team that would soon relocate to Hangzhou, where he learned that his assignment was to work on the OceanBase cloud platform, helping bring the OceanBase database service to Alibaba Cloud and serving as its internal operations platform.

The team was composed mostly of recent graduates, with only two more experienced members, including a veteran Alibaba employee. They began by establishing development processes and standards while trying to understand the product, which at the time lacked a clear definition, leading to frequent idea generation, rapid development, and continuous iteration.

The working environment was an open‑plan office in the Huanglong Times Square building, resembling a noisy internet café, where discussions were lively, debates intense, and the space buzzed with a “chaotic yet vibrant” energy that persisted late into the night.

After a few months, the product’s shape remained vague. Drawing on his prior experience with zero‑to‑one projects, the author quickly familiarized himself with databases, OceanBase, and its operations system, using intuition to judge which ideas were worth pursuing and which were not, while simultaneously studying the technology in his spare time.

He first tackled the team’s development workflow, standards, and tooling, which gradually improved the team’s awareness of best practices and boosted overall development efficiency. Together with his colleague, he began to systematically think about and define the product’s direction.

When the focus shifted to scaling OceanBase operations, real‑time monitoring, and cloud migration, the author chose to address the monitoring challenge first. The goal was to achieve second‑level real‑time monitoring for a cluster of thousands of machines, covering data and log collection, computation, and visualization. Because existing solutions like ELK did not fit the special business logic, they adopted JStorm, Alibaba’s internal stream‑processing system, to build a pipeline that collected KPI and log data, queued messages, performed aggregation and computation in JStorm, and persisted results for real‑time display. The system handled hundreds of metrics across thousands of machines, and continuous performance tuning made the pipeline increasingly stable.

In parallel, the author and a few teammates built a full‑link network monitoring solution for OceanBase nodes, capable of displaying topology and link quality in seconds. This capability proved critical when an international payment service experienced periodic jitter; the newly deployed network monitoring quickly identified the root cause, allowing the issue to be resolved.

During that year, the author maintained a weekly commute between Beijing and Hangzhou, lived in hotels with a predictable breakfast routine, and relied on his family’s support at home while balancing childcare responsibilities.

By the end of 2015, the OceanBase cloud product moved from a set of tools to a commercial offering, the team grew from a handful to nearly twenty members, and the author took charge of the product’s architecture, navigating Alibaba Cloud’s integration processes and completing the cloud migration before the fiscal year ended. He emphasizes the “speed‑first” philosophy he learned at Alibaba: think fast, act fast, validate fast, and iterate rapidly.

Reflecting on his first year, he notes a shift from disdain for “rough, fast, fierce” methods to embracing them, learning to handle uncertainty calmly, and adapting to a culture that values chaotic vitality over the soft‑skill‑focused environment of foreign companies. He concludes that despite pain and confusion, the experience taught him resilience and the importance of daring ideas and execution.

As the fiscal year closed, he began planning for the next year, only to encounter new changes before the roadmap was fully defined.