Table of contents

Managing Virtual Machines

Learn how to monitor, control, and maintain your virtual machines.

Viewing VMs

Navigate to Compute > Virtual Machines to see all VMs in your project.

VM List

The list displays:

  • Name: VM identifier
  • Status: Current state
  • OS: Operating system
  • Resources: CPU, memory, GPU allocation
  • IP Address: Public IP for SSH
  • Hourly Cost: Current billing rate

Filtering

  • Filter by status (Running, Stopped, All)
  • Search by VM name

VM Actions

Start a VM

For stopped VMs:

  1. Click the Start button
  2. Wait for status to change to Running (1-2 minutes)
  3. Billing resumes when running

Stop a VM

To pause a running VM:

  1. Click the Stop button
  2. Confirm the action
  3. VM enters Stopped state

When stopped:

  • CPU/GPU billing stops
  • Storage charges continue (reduced rate)
  • Data on disk is preserved
  • Public IP may be released

Restart a VM

To reboot:

  1. Click the Restart button
  2. VM will shutdown and start again

Useful for:

  • Applying kernel updates
  • Recovering from issues
  • Refreshing the system

Delete a VM

To permanently remove:

  1. Click Delete
  2. Confirm the deletion
  3. VM and all its data are removed

Warning: Deletion is permanent. Back up any important data first.

VM Details

Click on a VM to view:

Overview Tab

  • Full configuration
  • Public and private IPs
  • SSH connection string
  • Creation date
  • Associated project

Billing Tab

  • Current hourly rate
  • Total cost since creation
  • Breakdown by resource type

Monitoring

Resource Usage

VMs show basic metrics:

  • CPU utilization
  • Memory usage
  • Disk usage
  • Network traffic

For detailed monitoring, install tools inside the VM:

# Install htop for system monitoring
sudo apt install htop

# For GPU monitoring
nvidia-smi -l 1

GPU Status

For GPU VMs, verify GPU availability:

nvidia-smi

Check CUDA:

nvcc --version

Data Management

Backing Up Data

Before stopping or deleting a VM, back up important data:

Using SCP:

scp -r root@<vm-ip>:/path/to/data ./local_backup/

Using rsync:

rsync -avz root@<vm-ip>:/path/to/data/ ./local_backup/

Expanding Storage

To increase disk size:

  1. Stop the VM
  2. Contact support for disk expansion
  3. Start the VM
  4. Extend the filesystem inside the VM

Using NFS Volumes

For shared persistent storage:

  1. Create an NFS Volume
  2. Mount it inside your VM:
sudo mount -t nfs <nfs-server>:<share-path> /mnt/data

VM Snapshots

Snapshots capture the VM state for backup or cloning.

Creating a Snapshot

  1. Go to VM details
  2. Click Create Snapshot
  3. Enter a name for the snapshot
  4. Wait for completion

Managing Snapshots

View snapshots in VM Snapshots section:

  • List all snapshots
  • Restore from a snapshot
  • Delete old snapshots

Restoring from Snapshot

  1. Find the snapshot
  2. Click Restore
  3. A new VM is created from the snapshot

Best Practices

Cost Optimization

  • Stop VMs when not in use
  • Right-size resources based on actual usage
  • Delete VMs when projects are complete

Security

  • Keep the OS updated
  • Don’t expose unnecessary ports
  • Use SSH keys, not passwords
  • Consider firewall rules inside the VM

Reliability

  • Create regular snapshots
  • Use NFS for important data
  • Monitor disk space usage

Troubleshooting

Cannot Connect via SSH

  1. Verify VM is Running
  2. Check the correct IP address
  3. Ensure your SSH key matches
  4. Check if SSH service is running:
    # From VM console if available
sudo systemctl status sshd

VM is Slow

  • Check CPU/memory usage
  • Verify GPU is being utilized properly
  • Check for disk space issues
  • Review running processes

GPU Not Detected

# Check if driver is loaded
lsmod | grep nvidia

# Reinstall drivers if needed
sudo apt install nvidia-driver-535

Next Steps

Explore the GPU Marketplace for dedicated baremetal instances.