On Monday, China’s top cloud provider had a nearly two-hour-long disruption affecting customers in mainland China, H, and the United States. The impact was mainly felt by several of Alibaba’s cloud database management products, including its PostgreSQL, Redis, and MySQL editions. Beijing, Shanghai, Hong Kong, and Virginia in the U.S. were among eight regions that saw service affected. According to the firm, the problem was resolved by 10:58 on November 27. This is the second incident to hit Alibaba’s cloud platform within a month. It comes on the heels of a lengthy outage last December that impacted several of its customers’ platforms, including a crypto exchange and Macau’s Monetary Authority website.
The latest issue stemmed from what the company said was a refrigeration failure at one of its Hong Kong availability zones. The firm says the issues, which started at 5:44 pm local time, led to abnormalities in console and application programming interface access for certain services. It says the problems were resolved by 9:11 pm and that the actual operation of most Alibaba Cloud products remained unaffected.
Alibaba Cloud is a crucial computing infrastructure provider for the tech industry in China, where 80 percent of China’s tech companies and half of large AI model firms run their servers on the platform, CEO Eddie Wu Yongming said last week. It is a significant component of Alibaba’s broader push to position itself as an open-technology platform with the infrastructure services needed for AI innovation and transformation across “thousands of industries.”
Founded in 2011, Alibaba’s cloud division had revenues of US$12.5 billion in the first six months of this year. It is China’s most significant infrastructure as a service vendor, holding a 29.9 percent market share, ahead of Huawei and China Telecom. Besides its cloud business, Alibaba has a range of other digital offerings, including mobile messaging apps and e-commerce platforms.
The latest outage has pressured Alibaba to ensure it has robust systems, a transparent process for identifying the cause of issues, and better communication with customers when problems arise. It also highlights the importance of incorporating multiple layers of redundancy in cloud disaster recovery plans. The frequency of these outages is also likely to raise questions about the reliability of Alibaba Cloud for many businesses and users, especially those relying highly on its infrastructure. Preparing to list its cloud unit, the firm will be expected to deliver consistent uptime to its customers. A spokesman for Alibaba’s cloud division told the South China Morning Post that it has improved its monitoring and response mechanisms to prevent the recurrence of such incidents in the future. However, he didn’t respond to questions asking for further details about the root causes of Sunday’s disruption. The company is not the only public cloud provider to have experienced outages this month. Amazon’s AWS and Microsoft Azure have both had problems this month.