External Storage Access

Access and configure the 17.4TB external NVMe storage available on Voltage Park deployments.

Overview

Voltage Park nodes include substantial external storage in addition to the boot drive. While the main boot drive shows approximately 300GB, each node provides 6 additional NVMe drives totaling ~17.4TB of external storage capacity.

Storage Layout

nvme0n1 (~447GB) - Primary OS boot drive, mounted at /
nvme1n1 through nvme6n1 - 6 external data drives × 2.9TB each = ~17.4TB total

These external drives are available but not automatically mounted - you need to configure them based on your requirements.

Setup Options

Choose the configuration that best matches your needs:

Option 1: Individual Drives (Recommended)

Pros:

✅ If one drive fails, only data on that drive is lost
✅ Easy to recover - just mount the remaining drives
✅ Simple to manage and troubleshoot
✅ Can move drives to another system easily

Cons:

❌ Need to manage 6 separate mount points

Use Case: Best when data recovery and safety are priorities.

Option 2: RAID 6 (Balance of Capacity and Safety)

Pros:

✅ Survives up to 2 simultaneous drive failures
✅ Data remains accessible with failed drives
✅ Single large volume (~11.6TB usable)

Cons:

❌ Need RAID knowledge to recover
❌ Loses ~5.8TB to redundancy

Use Case: When you need both capacity and redundancy.

Option 3: RAID 0 (Maximum Capacity, NOT RECOMMENDED)

Pros:

✅ Full ~17.4TB usable capacity
✅ Single mount point

Cons:

❌ CRITICAL: If ANY drive fails, ALL data across ALL drives is lost
❌ No redundancy whatsoever

Use Case: Only for temporary/expendable data.

Option 4: LVM (Flexible Management)

Pros:

✅ Flexible volume management
✅ Can add/remove drives dynamically
✅ Single large volume

Cons:

❌ Without RAID, no redundancy (similar to RAID 0)
❌ More complex to manage

Use Case: When you need flexibility in storage management.

Recommended Setup: Individual Drives

This approach provides the best balance of simplicity and data safety.

Create Mount Points

sudo mkdir -p /mnt/nvme{1..6}

Format the Drives

Format each drive with ext4 filesystem:

sudo mkfs.ext4 /dev/nvme1n1
sudo mkfs.ext4 /dev/nvme2n1
sudo mkfs.ext4 /dev/nvme3n1
sudo mkfs.ext4 /dev/nvme4n1
sudo mkfs.ext4 /dev/nvme5n1
sudo mkfs.ext4 /dev/nvme6n1

Warning: This will erase all existing data on these drives.

Mount the Drives

for i in {1..6}; do 
  sudo mount /dev/nvme${i}n1 /mnt/nvme${i}
done

Make Mounts Persistent

Enable auto-mount on boot by adding entries to /etc/fstab:

for i in {1..6}; do 
  echo "/dev/nvme${i}n1 /mnt/nvme${i} ext4 defaults 0 2" | sudo tee -a /etc/fstab
done

Set Ownership

Make the drives writable by your user:

for i in {1..6}; do 
  sudo chown ubuntu:ubuntu /mnt/nvme${i}
done

Verify Setup

Check that all drives are mounted:

df -h | grep nvme

Expected output:

/dev/nvme0n1p2  439G   28G  389G   7% /
/dev/nvme0n1p1  511M  6.1M  505M   2% /boot/efi
/dev/nvme1n1    2.9T   28K  2.8T   1% /mnt/nvme1
/dev/nvme2n1    2.9T   28K  2.8T   1% /mnt/nvme2
/dev/nvme3n1    2.9T   28K  2.8T   1% /mnt/nvme3
/dev/nvme4n1    2.9T   28K  2.8T   1% /mnt/nvme4
/dev/nvme5n1    2.9T   28K  2.8T   1% /mnt/nvme5
/dev/nvme6n1    2.9T   28K  2.8T   1% /mnt/nvme6

Usage

Accessing the Drives

Each drive is accessible at its mount point:

# Access drive 1
cd /mnt/nvme1
 
# Create files
echo "test data" > /mnt/nvme1/myfile.txt
 
# List contents
ls -lh /mnt/nvme1/

Checking Space Usage

# Check space on all drives
df -h | grep nvme
 
# Check space on a specific drive
df -h /mnt/nvme1

Distributing Data

You can distribute your data across all drives:

# Store different datasets on different drives
cp -r /path/to/dataset1 /mnt/nvme1/
cp -r /path/to/dataset2 /mnt/nvme2/
# ... and so on

Alternative Setup: RAID 6

If you prefer redundancy over maximum capacity:

Install RAID Tools

sudo apt update
sudo apt install mdadm

Create RAID 6 Array

# Create RAID 6 (survives 2 drive failures)
sudo mdadm --create /dev/md0 --level=6 --raid-devices=6 \
  /dev/nvme1n1 /dev/nvme2n1 /dev/nvme3n1 \
  /dev/nvme4n1 /dev/nvme5n1 /dev/nvme6n1
 
# Format the array
sudo mkfs.ext4 /dev/md0
 
# Create mount point and mount
sudo mkdir -p /mnt/raid
sudo mount /dev/md0 /mnt/raid
 
# Make persistent
echo '/dev/md0 /mnt/raid ext4 defaults 0 2' | sudo tee -a /etc/fstab
 
# Save RAID configuration
sudo mdadm --detail --scan | sudo tee -a /etc/mdadm/mdadm.conf
sudo update-initramfs -u

This provides ~11.6TB usable space with 2-drive fault tolerance.

Alternative Setup: LVM

For maximum flexibility in storage management:

Install LVM Tools

sudo apt update
sudo apt install lvm2

Create LVM Setup

# Create physical volumes
sudo pvcreate /dev/nvme{1..6}n1
 
# Create volume group
sudo vgcreate data_vg /dev/nvme{1..6}n1
 
# Create logical volume with all space
sudo lvcreate -l 100%FREE -n data_lv data_vg
 
# Format and mount
sudo mkfs.ext4 /dev/data_vg/data_lv
sudo mkdir -p /mnt/data
sudo mount /dev/data_vg/data_lv /mnt/data
 
# Make persistent
echo '/dev/data_vg/data_lv /mnt/data ext4 defaults 0 2' | sudo tee -a /etc/fstab

Troubleshooting

Checking Drive Status

# List all block devices
lsblk
 
# Check drive health
sudo smartctl -a /dev/nvme1n1  # Requires smartmontools package

Drives Don't Mount on Boot

# Check fstab syntax
cat /etc/fstab
 
# Try manual mount to test
sudo mount -a
 
# Check system logs
sudo journalctl -xe | grep mount

Unmounting Drives

# Unmount a specific drive
sudo umount /mnt/nvme1
 
# Unmount all data drives
for i in {1..6}; do 
  sudo umount /mnt/nvme${i}
done

Removing Drives from fstab

To undo the auto-mount configuration:

# Edit fstab and remove the nvme entries
sudo nano /etc/fstab
 
# Or use sed to remove them
sudo sed -i '/\/mnt\/nvme[1-6]/d' /etc/fstab

Recovery Scenarios

If One Drive Fails

With individual drives setup:

Identify the failed drive using dmesg or lsblk
The other 5 drives remain fully accessible
Only data on the failed drive is lost
Replace the failed drive and format it
Restore data from backups for that drive only

Moving Drives to Another System

Unmount the drives: sudo umount /mnt/nvme{1..6}
Physically move the drives
On the new system, mount them: sudo mount /dev/nvmeXn1 /mnt/target
All data remains intact

Best Practices

Organization:

Keep track of what data is stored on which drive

Use descriptive names via symlinks:

ln -s /mnt/nvme1 /mnt/datasets
ln -s /mnt/nvme2 /mnt/models

Monitoring:

Install and use smartmontools to monitor drive health
Check space regularly to avoid running out: df -h

Data Safety:

Maintain regular backups of critical data
Even with redundancy, backups are essential
Test your backup restoration process

Planning:

Plan your data distribution strategy before filling drives
Document which data goes where
Consider future scalability needs

Summary

Default Configuration:

Boot drive shows ~300GB (nvme0n1)
6 external drives available but not mounted by default
Each external drive provides ~2.8TB
Total external capacity: ~17.4TB

Recommended Approach:

Mount drives individually at /mnt/nvme1 through /mnt/nvme6
Enable auto-mount for convenience
Distribute data across drives based on project needs
Maintain backups of critical data

Why Individual Drives:

Maximum data recoverability
Simple management
One drive failure doesn't affect others
Easy to understand and troubleshoot

Additional Resources

SSH Connection Setup - Secure instance access
Getting Started - Complete setup guide
Security - Best practices for secure storage