Alibaba Group Holding’s cloud computing arm is now home to 500,000 terabytes worth of data from Chinese lifestyle platform Xiaohongshu after what the companies called the largest data migration ever, a case that could enhance the leading position of one of the country’s largest tech firms in the domestic cloud market.

The migration of the 500-petabyte “data lake” – a repository that stores, processes and secures large amounts of structured and unstructured data – started last November, taking a year of 1,500 staff members from Xiaohongshu working with teams at Alibaba to complete, according to a statement from Alibaba Cloud. Alibaba owns the South China Morning Post.

The data lake contains all the raw and critical data that Xiaohongshu has accumulated since it started operating 11 years ago. For context, one petabyte of data can hold around 11,000 4K high-definition movies, which would require more than two-and-a-half years of non-stop binge watching to get through – assuming an average size of 90 gigabytes and length of two hours per film.

The case marks a big win for Alibaba as cloud providers are racing to upgrade their data centres amid rising demand for artificial intelligence services. Alibaba is the largest cloud provider in China with 36 per cent of the market in the second quarter, according to Canalys. It is also the largest cloud provider in Asia-Pacific and third-largest in the world by revenue.

Shanghai-based Xiaohongshu is the largest lifestyle social media platform in China, boasting a monthly active user base of more than 300 million and growing rapidly in recent years. The vast amount of data it generates daily added to migration challenges that were “beyond one’s imagination”, Alibaba Cloud said in its statement posted to WeChat.

The scale of this project could bolster confidence in Alibaba Cloud after outages in recent years have dented its image. Outages in December 2022 and November 2023 took websites offline in Hong Kong and Macau, and suspended service on TaobaoAlibaba’s flagship shopping platform.