Cloudfoxy Labs

← All labs

Lab 10: Docker and Container Toolkit break-and-fix

Level 4 (Troubleshoot) · GPU tier

150 minutes suggested, up to 300 in one session. Price: 5 credits per minute. Includes Check my work.

Hosted, this lab covers Parts A, B, C, D. The other Parts need your own machine (Home Lab Kit) and are shown collapsed in the lab.

The GPU tier is not available on this site yet.

What you will be able to do

  • Name the piece of the container stack that each GPU request needs: --gpus, --runtime=nvidia and --device nvidia.com/gpu=all.
  • Read docker info, journalctl -u docker and dockerd --validate to find out what the daemon actually loaded.
  • Explain why dockerd must be restarted after you install the toolkit or edit daemon.json, and predict how Docker 28 and Docker 29.2 or later differ.
  • Repair a broken daemon.json, a missing runtime entry, a missing CDI specification and a missing group membership, each so that it survives a reboot.
  • Reconstruct Chapter 18's real incident: a package purge that deleted daemon.json and an autoremove that took the toolkit with it.
  • Tell the two NVML failures apart: driver/library mismatch on the host and lost device access inside a container.