Kubernetes, the market-leading container orchestration platform, is celebrating its 10th anniversary in mid-2024. Xing Yang, the cloud-native storage tech lead at VMware by Broadcom, began working on storage in Kubernetes in 2017, focusing on custom resource definitions (CRDs) to enhance the platform’s extensible core.
As Kubernetes gained market leadership, Yang worked on container storage interface (CSI) and Kubernetes Operators, based on CRDs. These additions brought storage and data protection functionality while maintaining Kubernetes’ core characteristics. To mark Kubernetes’ first decade, a series of interviews with engineers involved in developing the platform and addressing storage challenges, including the use of Kubernetes Operators, have been planned as they look ahead to a future driven by artificial intelligence (AI) workloads.
When Kubernetes first launched, the container orchestration market was still emerging, with Docker gaining popularity for building images. As Kubernetes evolved to deploy Docker images on distributed systems, it became the preferred container orchestration system.
Yang’s involvement in Kubernetes began in 2017, contributing to the VolumeSnapshot project in Kubernetes SIG Storage. In collaboration with the community, Yang championed the inclusion of advanced storage functionalities like Volume Snapshot, Volume Cloning, and CSI topology to support the growing number of stateful workloads running in Kubernetes.
As Kubernetes matured, Yang focused on Kubernetes Operators, collaborating with the Data on Kubernetes Community to develop white papers on running databases in Kubernetes. Operators, based on CRDs, automate Day Two operations for databases, making deployment, monitoring, and lifecycle management more efficient in a cloud-native environment.
Looking back on Kubernetes’ journey over the past decade, Yang acknowledges the progress made in supporting data workloads and the platform’s maturity. However, challenges remain around Day Two operations, Kubernetes complexity, and the lack of standardization for Operators in multi-cluster environments.
Despite the challenges, Yang remains optimistic about Kubernetes’ future and its potential to innovate in the coming years.