Skip to main content

Troubleshooting

Common issues you may encounter when working with Ephemeral, Persistent, or Shared Disks and how to resolve them.

Resized a Persistent Disk but the filesystem did not grow​

Cause: Resizing a Persistent Disk changes the underlying block device size, but does not grow the filesystem inside the VM. The filesystem must be extended manually after the disk is re-attached.

Solution: Follow the steps in Growing the filesystem after a resize.

vastnfs-ctl: command not found​

Cause: The VAST NFS driver required for mounting Shared Disks is not installed on the VM.

Solution: Install the driver using the steps in Setting up the VAST NFS driver.

umount: target is busy when unmounting a Shared Disk​

Cause: A process on the VM has an open file handle or working directory on the mount path, so the kernel cannot release the mount.

Solution: Find and stop the processes holding the mount:

sudo lsof +f -- /path/to/mount
sudo fuser -m /path/to/mount

Then retry sudo umount /path/to/mount. If the processes cannot be stopped, sudo umount -l /path/to/mount performs a lazy unmount that completes when all handles are closed.

VM fails to boot after editing /etc/fstab​

Cause: A malformed or unresolvable entry in /etc/fstab blocks the boot process while waiting for the mount.

Solution: Use serial console access to drop into single-user mode and restore /etc/fstab from a backup, or comment out the offending line. To reduce the risk of boot failures, include the nofail and x-systemd.automount options on Shared Disk entries — see the fstab example.

Persistent Disk does not appear in lsblk after attaching​

Cause: The VM kernel has not yet picked up the newly attached device, or the attach operation is still in progress on the backend.

Solution:

  1. Confirm the disk is attached in the Crusoe Cloud console or via crusoe storage disks list.

  2. Run lsblk again after a few seconds.

  3. If the disk is still missing, rescan the virtio bus:

    echo 1 | sudo tee /sys/bus/pci/rescan

  4. If the disk remains invisible, detach and re-attach the disk via the console, CLI, or Terraform.

VM crashed after detaching a Persistent Disk​

Cause: The disk was detached while still mounted. Detaching a mounted Persistent Disk can cause the VM kernel to panic and, in some cases, results in loss of data on both the persistent and ephemeral disks.

Solution: Always unmount the disk before detaching:

sudo umount /dev/vdX

See Detaching Persistent Disks.