Skip to content

Sesterce: mounting shared storage

Mount persistent block storage volumes on Sesterce instances using ext4-formatted block devices.

Overview

Sesterce persistent volumes attach as raw block devices (e.g., /dev/vdb) and format with the ext4 filesystem. A stable UUID identifies each volume, referenced in /etc/fstab for persistent mounting.

Protocol: Block device (ext4) Connection: Directly attached block device Performance: Local disk throughput

Prerequisites

Before starting, ensure you have:

  1. Created a volume with provider: sesterce via the Spheron dashboard (Create Volume page)
  2. Attached the volume to your Sesterce instance
  3. SSH access to your Sesterce instance (credentials in the instance's Details drawer)

Mounting process

Connect to your instance

SSH into your Sesterce GPU instance using the connection details from the instance's Details drawer:

ssh ubuntu@<your-instance-ip>

See SSH Connection Setup for detailed connection instructions.

Identify the new disk

List all block devices to confirm the volume is visible and unmounted:

lsblk -o NAME,SIZE,TYPE,FSTYPE,MOUNTPOINT

Expected output:

NAME     SIZE TYPE FSTYPE   MOUNTPOINT
vda      100G disk
├─vda1  99.9G part ext4     /
vdb       50G disk                   # new disk, no MOUNTPOINT

If the disk has no FSTYPE and no MOUNTPOINT, it is raw and ready to format.

Format the disk

Format the volume with the ext4 filesystem:

sudo mkfs.ext4 /dev/vdb

Create mount directory

Create a directory where you'll mount the volume. Replace <VOLUME_NAME> with your preferred directory name:

sudo mkdir -p /mnt/<VOLUME_NAME>

Common choices include:

  • /mnt/data: simple and clear
  • /mnt/storage: standard mount location
  • /mnt/datasets: for ML/AI datasets
  • /mnt/models: for model checkpoints

Mount the disk

Mount the formatted disk to the mount point:

sudo mount /dev/vdb /mnt/<VOLUME_NAME>

Verify it is now visible:

df -h

Look for your volume in the output:

Filesystem       Size  Used Avail Use% Mounted on
...
/dev/vdb          49G   24K   47G   1% /mnt/data

Persist the mount in /etc/fstab (recommended)

Device names like /dev/vdb can change across reboots. Use the disk's UUID for a stable reference.

Get the UUID:

sudo blkid /dev/vdb

Example output:

/dev/vdb: UUID="a1b2c3d4-e5f6-7890-abcd-ef1234567890" TYPE="ext4"

Add the entry to /etc/fstab (replace the UUID with your own):

echo 'UUID=<your-uuid>  /mnt/<VOLUME_NAME>  ext4  defaults  0  2' | sudo tee -a /etc/fstab
Field explanation:
FieldValueMeaning
deviceUUID=...Identifies the disk by UUID
mount point/mnt/<VOLUME_NAME>Where the disk is accessible
typeext4Filesystem type
optionsdefaultsStandard mount options
dump0Disable backup via dump
pass2fsck order (2 = non-root disk)

Verify the fstab entry works:

sudo mount -a

No errors confirms the entry is correct.

Set permissions (optional)

Make the mounted storage writable by your user:

sudo chown sesterce:sesterce /mnt/<VOLUME_NAME>

Replace sesterce:sesterce with your username if different.

Use your mounted volume

Access the storage

Once mounted, use the volume like any local directory:

# Navigate to the storage
cd /mnt/data
 
# Create files and directories
mkdir my-project
echo "Hello, storage!" > my-project/readme.txt
 
# List contents
ls -lh /mnt/data/

Check storage usage

# Check space on the mounted volume
df -h /mnt/data
 
# Check detailed disk usage
du -sh /mnt/data/*

Work with large datasets

Download datasets and checkpoints to the mounted volume. Data persists across instance restarts.

# Example: Download dataset to storage
cd /mnt/data
wget https://example.com/large-dataset.tar.gz
tar -xzf large-dataset.tar.gz

Unmount shared storage

Unmount when swapping volumes, detaching the volume, or terminating the instance.

Unmount the volume

sudo umount /mnt/<VOLUME_NAME>

If you get a "target is busy" error, check for active processes:

# Check what's using the storage
lsof /mnt/<VOLUME_NAME>
 
# Or use fuser
fuser -m /mnt/<VOLUME_NAME>

Remove from file system table

Remove the entry from /etc/fstab to prevent auto-mount on next boot:

# Edit fstab manually
sudo nano /etc/fstab
 
# Or remove automatically with sed
sudo sed -i '/UUID=<your-uuid>/d' /etc/fstab

Troubleshooting

Disk not visible in lsblk

Cause: Volume not attached by the cloud provider.

Solution:
  1. Check the Spheron AI dashboard and confirm the volume is attached to your instance
  2. Re-attach the volume if needed and reconnect to the instance

mkfs.ext4 fails

Cause: Disk is already mounted.

Solution: 
sudo umount /dev/vdb
sudo mkfs.ext4 /dev/vdb

Volume not mounting on boot

Cause: Missing or incorrect /etc/fstab entry.

Solution: Verify the fstab entry exists and the UUID matches:

sudo blkid /dev/vdb
cat /etc/fstab
sudo mount -fav

Permission denied when writing to disk

Cause: Mount point ownership is root.

Solution: 
sudo chown sesterce:sesterce /mnt/<VOLUME_NAME>

Check mount status

# List all mounted filesystems
mount | grep /mnt
 
# Check filesystem integrity (disk must be unmounted)
sudo fsck /dev/vdb
 
# Verify fstab syntax
sudo mount -fav

Best practices

Organization:
  • Use descriptive mount points: /mnt/data, /mnt/models, /mnt/datasets
  • Create subdirectories for different projects
  • Document what data is stored where
Performance:
  • Block storage provides local disk throughput; no network overhead
  • Use ext4 for general workloads; consider xfs for large file performance
Data safety:
  • Maintain backups of critical data
  • Volumes persist independently of instances, but data can still be lost due to accidental deletion
  • Test backup and restore procedures
Security:
  • Restrict access using filesystem permissions
  • Audit who has access to mounted storage

What's next

Powered by Spheron
© 2026 Spheron Network. All rights reserved.