Skip to main content

Product Introduction

What is a Volume?

Overview

A Volume is a storage resource provided by the Bitdeer AI platform that allows users to persist data across different services and workflows. Volumes are designed to store and retain data independently of compute instances under AI Studio, such as: (1) Notebooks, (2) Containers, and (3) Training jobs. This ensures seamless data accessibility and management.

Volumes are particularly useful for:

  • Data Persistence: Retain data across sessions, even after compute resources are terminated.
  • Shared Storage: Share data among multiple compute instances or users.
  • Scalable Storage: Accommodate varying storage needs by customizing volume size and access modes.

Key Features

  1. Persistent Storage:
    • Volumes ensure that your data remains intact even after the associated compute instances (e.g., notebooks, containers) are stopped or deleted.
  2. Flexible Configurations:
    • Customize volume size, storage class (e.g., standard or high-performance), and access modes to suit your workload.
  3. Integration Across Services:
    • Volumes can be created and managed independently or as part of creating notebooks, containers, and training jobs.
  4. Ease of Management:
    • Manage volumes directly from the AI Studio Console, with options to create, list, and delete volumes through an intuitive user interface.

Typical Use Cases

  1. Data Sharing:
    • Use a shared volume to store datasets, models, or logs that can be accessed by multiple notebooks or containers.
  2. Data Backup and Recovery:
    • Save important data in a persistent volume to prevent loss due to compute instance termination.
  3. Scalable Machine Learning Workflows:
    • Attach volumes to training jobs or inference services to ensure efficient data handling across workflows.

How Volumes Work

Volumes operate independently of compute resources. They are attached to notebooks, containers, or training jobs as needed, providing a seamless and persistent storage experience. When the associated compute resource is terminated, the volume remains available for reuse or deletion.

Managing Volumes

  1. Creating Volumes:
    • Volumes can be created through the Volume Management Dashboard or during the creation of compute instances such as notebooks and containers.
  2. Listing Volumes:
    • View all existing volumes, along with their details, such as size, storage class, and access modes.
  3. Deleting Volumes:
    • Once a volume is no longer needed, it can be deleted to free up storage resources.

For more details, refer to the Volume Management User Guide.