runai-bgu projects Manual

Introduction

runai-bgu projects is a command-line interface (CLI) for managing projects on the BGU HPC cluster. This command provides access to project information and allows you to view all available projects you have access to. Projects are organizational units that group workloads and manage resource allocation on the cluster.

This manual explains how to use runai-bgu projects to view and manage your project access.

Quick Start

To list all available projects, use:

$ runai-bgu projects

Shows all projects you have access to on the cluster.

Basic Usage

List All Projects

Show all projects available to you:

$ runai-bgu projects

Displays all projects you can access, including project names, descriptions, and your access level.

Understanding Projects

Projects serve several important functions:

Resource Organization: Projects group related workloads and provide resource boundaries for teams or research groups.
Access Control: Projects control who can submit workloads and access specific cluster resources.
Resource Allocation: Each project has allocated quotas for CPU, memory, GPU, and storage resources.
Billing and Accounting: Resource usage is tracked and reported at the project level for accounting purposes.

Project Information

When you list projects, you’ll see:

Project Name: The unique identifier for the project, used when submitting workloads.
Description: A human-readable description of the project’s purpose.
Access Level: Your permission level within the project (viewer, editor, admin).
Resource Quotas: The allocated limits for various resources within the project.
Current Usage: How much of the allocated resources are currently in use.

Common Use Cases

Check Available Projects

See which projects you can submit workloads to:

$ runai-bgu projects

Find Project for Workload Submission

When submitting workloads, you need to know which projects are available:

$ runai-bgu projects
$ runai-bgu submit python -n my-job -c 2 -m 4Gi -p project-name

Project Management

While runai-bgu projects focuses on listing projects, remember that:

Default Project: You can set a default project using runai-bgu config <project-name> to avoid specifying -p for every command.
Project Selection: When submitting workloads, you can specify a project with the -p flag if it’s different from your default.
Resource Limits: Each project has resource quotas that limit how many resources you can use simultaneously.