runai-bgu list Manual

Introduction

runai-bgu list is a command-line interface (CLI) for listing workloads on the BGU HPC cluster. This command provides an overview of all your workloads including both workspace and training jobs. The command automatically shows workloads from your default project unless specified otherwise, and supports various filtering and output formatting options.

This manual explains how to use runai-bgu list to view and manage your workloads.

Quick Start

To list all your workloads, use:

$ runai-bgu list

Lists all workloads in your default project.

Basic Usage

List All Workloads

Show all workloads in your default project:

$ runai-bgu list

Displays all workspace and training workloads in your default project with basic information like name, status, and creation time.

List Workloads from All Projects

Show workloads from all projects you have access to:

$ runai-bgu list -A

The -A flag lists workloads from all projects, not just your default project.

Filter by Status

Show only workloads with a specific status:

$ runai-bgu list --status Running

Filters to show only workloads that are currently in "Running" status.

Output Formats

JSON Output

Get workload information in JSON format for programmatic use:

$ runai-bgu list --json

Outputs the workload list in structured JSON format, useful for scripts and automation.

YAML Output

Get workload information in YAML format:

$ runai-bgu list --yaml (1)

1	Outputs the workload list in YAML format for human-readable structured data.

Table Without Headers

Display results in table format without column headers:

$ runai-bgu list --no-headers (1)

1	Shows the table output without header row, useful for parsing in scripts.

Pagination Options

Limit Results

Control the maximum number of items returned:

$ runai-bgu list --max-items 50

Limits the output to maximum 50 workloads.

Page Size

Set the page size for paginated results:

$ runai-bgu list --page-size 20

Sets each page to contain 20 workloads (default is 100).

Disable Pagination

Get all results in a single page:

$ runai-bgu list --no-pagination

Returns all results without pagination, which may be slower for large numbers of workloads.

Project Selection

List workloads from a specific project:

$ runai-bgu list -p my-research-project

Shows workloads only from the specified project "my-research-project".

Understanding the Output

The list command provides essential information about your workloads:

Name: The unique identifier for each workload.
Type: Whether it’s a workspace (interactive) or training (batch) workload.
Status: Current state of the workload (Running, Pending, Completed, Failed, etc.).
Project: The project the workload belongs to.
Age: How long ago the workload was created.
Resources: Summary of allocated CPU, memory, and GPU resources.