Managing Resources

Monitor and manage GPU instances and storage volumes.

Resources Page

Access via Resources in the sidebar. View three tabs:

• On-demand GPUs: Active and transitioning instances
• Volumes: Storage volumes
• Deleted: Terminated instances history

Instance Information

For each instance, view:

• Instance name
• Status (active, creating, pending, etc.)
• GPU type and count
• Region
• Hourly cost
• Created date
• IP address (when active)
• SSH connection details
• Attached volumes
• Tags

Instance Status

Status	Description	Billing
creating	Provisioning infrastructure	No
pending_provider	Cloud provider setup	No
pending	Final configuration	No
active	Running and accessible	Yes
failed	Launch failed	No (refunded)
suspended	Insufficient balance	Paused
deleting	Terminating	No
deleted	Removed	No

Terminate Instance

To stop billing and remove an instance:

1. Locate instance on Resources page
2. Click Delete/Terminate
3. Confirm deletion

What happens:

• Status changes to deleting
• Billing stops immediately
• Instance shuts down (1-2 minutes)
• Status changes to deleted
• Attached volumes are safely detached
• Instance moves to Deleted tab

Important: Instance storage is permanently deleted. Attached volumes are preserved.

Volume Management

View Volumes

The Volumes tab shows:

• Volume name
• Size (GB)
• Status (active, attached, deleting)
• Hourly cost
• Region
• Attached instance (if any)
• Created date

Volume Status

• active: Available, can be attached
• attached: Connected to instance
• deleting: Being removed
• suspended: Insufficient balance

Create Volume

1. Click "Create Volume"
2. Enter name (descriptive identifier)
3. Select region (must match future instance region)
4. Choose size (50 GB to 10,000 GB)
5. Review pricing
6. Click Create

Volume is billed hourly from creation, even when not attached.

Delete Volume

Warning: Permanently deletes all data.

1. Ensure volume is detached (status: active)
2. Click Delete
3. Confirm by typing volume name
4. Volume removed, billing stops

Cost Monitoring

Per-instance costs displayed:

• Hourly rate
• Running time
• Total cost accumulated

Wallet balance:

• View current balance
• Track spending in transaction history

Cost optimization:

• Terminate idle instances
• Delete unused volumes
• Use startup scripts to auto-shutdown after jobs
• Right-size GPU selection

Connecting to Instances

When status is active, use provided connection details:

ssh -i /path/to/key.pem root@<IP_ADDRESS>

Monitor GPU:

nvidia-smi
watch -n 1 nvidia-smi  # Update every second

Monitor system:

htop  # CPU and memory
df -h  # Disk usage

Deleted Resources

View history of terminated instances in the Deleted tab.

Information retained:

• Instance name and specifications
• Total cost incurred
• Created and deleted dates
• Deletion reason (if failed)

Deleted instances retained for 90 days, then permanently removed. Transaction history retained indefinitely.

Common Issues

Cannot connect via SSH:

• Verify status is active
• Confirm correct IP address
• Check key permissions: chmod 600 key.pem
• Wait 1-2 minutes after status becomes active

Instance suspended:

• Add funds to wallet
• Instance resumes automatically when balance sufficient

Volume not showing:

• Check region matches instance region
• Verify volume status is active
• Ensure not attached to another instance

Unexpected charges:

• Review active instances on Resources page
• Check for unattached volumes (still billing)
• Review transaction history for details