Skip to main content
Version: 0.0.1

Team Leader Guide

This guide is for team leaders who manage projects, monitor budgets, and allocate resources within their team.

Responsibilities​

As a team leader, you are responsible for:

  • ✅ Creating and managing projects (namespaces)
  • ✅ Allocating quotas to projects
  • ✅ Monitoring team balance and consumption
  • ✅ Requesting recharges when needed

Getting Started​

1. Access Bison​

Log in to the Web UI with your credentials.

2. View Team Dashboard​

Your dashboard shows:

  • Team balance and status
  • Resource utilization
  • Active projects
  • Cost trends

Managing Projects​

Create a Project​

  1. Navigate to Projects page
  2. Click Create Project
  3. Fill in:
    • Project Name: e.g., "training-ml-models"
    • Description: Project purpose
    • Quota (optional):
      • CPU: e.g., "8" cores
      • Memory: e.g., "32Gi"
      • GPU: e.g., "2"
  4. Click Create

List Projects​

# Via kubectl (if you have access)
kubectl get namespaces -l capsule.clastix.io/tenant=your-team

# Via API
curl http://localhost:8080/api/v1/teams/your-team/projects

Delete a Project​

  1. Navigate to Projects page
  2. Click Delete on the project row
  3. Confirm deletion

Warning: This will delete all resources in the project!

Monitoring Budget​

Check Balance​

View your current balance:

  1. Navigate to Team page
  2. See balance in the status card

Analyze spending patterns:

  1. Navigate to Reports page
  2. Select time range (7 days, 30 days, 90 days)
  3. View:
    • Cost breakdown by resource type
    • Daily cost trends
    • Per-project consumption

Request Recharge​

When balance is low:

  1. Click Request Recharge button
  2. Enter requested amount
  3. Add justification
  4. Submit request to administrator

Resource Management​

Monitor Quota Usage​

Check how much of your quota is being used:

kubectl describe quota -n your-project

Optimize Costs​

Tips to reduce spending:

  • Right-size resources: Don't over-provision CPU/Memory
  • Clean up idle pods: Delete unused workloads
  • Use spot/preemptible instances: Where applicable
  • Monitor GPU utilization: Ensure GPUs are fully utilized

Best Practices​

Project Organization​

  • Create separate projects for different workloads
  • Example: ml-training, ml-inference, data-processing

Quota Allocation​

  • Allocate quotas based on project priority
  • Reserve buffer for urgent tasks

Cost Awareness​

  • Review costs weekly
  • Identify and eliminate waste
  • Set up cost alerts

Next Steps​