Documentation

Scheduled Termination

The Scheduled Termination feature in VoltageGPU allows you to automatically terminate your running pods at a future date and time. Control costs and manage pod lifecycles without manual monitoring.

Overview

When you schedule a termination, the system automatically removes your pod at the specified time, helping you optimize resource usage and control your GPU rental costs more effectively.

Cost Control

Prevent unexpected charges by setting automatic termination times.

Set & Forget

No need to manually monitor and stop pods after your work is complete.

Flexible Scheduling

Choose preset intervals or set a custom date and time.

Cancellable

Change your mind? Cancel scheduled termination anytime before execution.

How to Schedule Pod Termination

Method 1: Once Pod is Deployed

If you already have a running pod and want to schedule its termination:

  1. Navigate to your pod detail page in the VoltageGPU dashboard
  2. Click the "DELETE" button in the top right corner
  3. Select "Schedule pod deletion" in the popup menu
  4. Specify the termination time by choosing from:
    • Preset time intervals: 1 hour, 2 hours, 3 hours, 6 hours, 12 hours, or 1 day
    • Custom date and time: Select a specific date and time for termination
💡 Tip

Use preset intervals for quick scheduling, or custom date/time for precise control over when your pod terminates.

Method 2: When Deploying a Pod

If you want to set auto-termination while creating a new pod:

  1. During the pod deployment process, locate the "Auto-Termination (Optional)" field
  2. Fill in the number of hours after which the pod should automatically terminate
  3. Complete the rest of the deployment configuration
  4. Deploy the pod - it will automatically terminate after the specified duration
✅ Pro Tip

Setting auto-termination during deployment is perfect for time-limited experiments, overnight training runs, or when you know exactly how long you need the GPU.

Managing Scheduled Termination

View Scheduled Termination Time

The scheduled termination time is displayed in the pod detail page. You can see exactly when your pod is set to terminate.

Cancel Scheduled Termination

If you change your mind and want to keep your pod running:

  1. Navigate to your pod detail page
  2. Click the "Remove Scheduled Terminate" button
  3. The scheduled termination will be cancelled
⚠️ Important

You can only cancel a scheduled termination before it executes. Once the termination time is reached, the pod will be deleted and cannot be recovered.

Common Use Cases

Overnight Training

Schedule termination for the morning after starting an overnight training run. Your pod stops automatically when training should be complete.

Budget Control

Set a maximum runtime to ensure you don't exceed your budget. Perfect for experiments with uncertain completion times.

Time-Limited Access

Provide temporary GPU access to team members with automatic cleanup. No need to remember to delete pods later.

Batch Processing

Run batch jobs with known durations and automatically free up resources when processing is complete.

Billing

  • Billing stops immediately when the pod is terminated at the scheduled time
  • You are only charged for the time the pod was running
  • No additional charges for using the scheduled termination feature
💰 Cost Savings

Scheduled termination helps prevent "forgotten pods" that continue running and incurring charges after your work is complete.

Frequently Asked Questions

Can I cancel a scheduled termination?

Yes, you can cancel the scheduled termination before it executes by clicking the "Remove Scheduled Terminate" button in the Pod detail page.

Will I be charged after the scheduled termination?

No, billing stops immediately when the pod is terminated at the scheduled time.

Can I schedule termination for multiple pods at once?

Currently, termination must be scheduled individually for each pod through its detail page.

What happens to my data when the pod terminates?

Local storage is deleted with the pod. Make sure to save important data to volumes or configure backups before termination.

Can I see when my pod is scheduled to terminate?

Yes, the scheduled termination time is displayed in the pod detail page.

What if I need more time than originally scheduled?

Cancel the scheduled termination before it executes, then optionally schedule a new termination time if needed.

Best Practices

  • Always configure backups before scheduling termination to protect your data
  • Add buffer time to your estimated runtime to account for unexpected delays
  • Save checkpoints regularly during long training runs
  • Use volumes for data that needs to persist beyond pod termination
  • Set up notifications to be alerted before termination (if available)

Ready to Optimize Your GPU Costs?

Schedule automatic termination and never pay for idle pods again.