• Google Kubernetes Engine (GKE) has seen a significant increase in the number of Kubernetes clusters running stateful applications since 2019.
• To cater to the rise in AI/ML workloads and data on Kubernetes, Google has introduced Cloud Storage FUSE for machine learning.
• With this, objects in Cloud Storage buckets can be accessed as files mounted as a local file system offering a frictionless experience for applications.
• The new Cloud Storage FUSE Container Storage Interface (CSI) driver on GKE allows Kubernetes applications to mount cloud storage buckets as local file systems, providing portability, massive scale, and built-in support for GKE Standard and Autopilot.
• The Cloud Storage FUSE CSI driver eliminates the need for privileged access and supports all accelerators available on GKE.
• This feature enables ML training, inference models, reading and checkpointing, and Python-based third-party data apps, among others.