Skip to Content
Scheduler AdministrationOverview

Scheduler Administration

The Fuzzball scheduler is responsible for placing and managing jobs on compute nodes in your cluster. As a cluster administrator, you can monitor and manage scheduler behavior through various administrative commands.

Prerequisites

All scheduler administration operations require that a fuzzball admin context is set up and authorized like so:

$ fuzzball context create <context_name> <api_url>

$ fuzzball context login -u <admin_username> -p '<admin_password>'

Common Administrative Tasks

Preparing for Maintenance

  1. Cordon nodes to prevent new workflow placement
    1. Allows running workflows to complete
  2. Perform maintenance operations
  3. Uncordon nodes to resume normal scheduling

Capacity Planning

  • Monitor node utilization and availability
  • Plan for cluster scaling operations
  • Manage resource allocation across nodes
TroubleshootingOverview