TroubleByteSearch
SOURCE VERIFIEDSource review Sep 8, 2026No hands-on test claimed2 sources

Docker: no space left on device

Docker can consume substantial storage through images, stopped containers, build cache and volumes. A no-space error may also come from the host filesystem or inode exhaustion. Docker provides docker system df to measure daemon usage and conservative prune commands that only remove categories you explicitly choose.

AuthorshipWritten and source-reviewed by Mihailo Ivanjac
Record created Sep 8, 2026Last source review Sep 8, 2026Source review only
SymptomDocker pulls, builds or container writes fail with no space left on device, or the Docker data filesystem is full/near full.
Problem familyServers · Docker
Highest step riskCaution
Editorial statePublic launch · indexed
Affected scope
Docker EngineLinux hostsImages and build cacheContainers and volumes
Start here

Measure before deleting

Run host disk/inode checks plus docker system df before deleting anything. Prune only unused Docker objects you understand; volumes are intentionally not removed by default because they may contain important data.

  1. Check df -h and df -i for the Docker data filesystem.
  2. Run docker system df and optionally docker system df -v.
  3. Identify reclaimable images, stopped containers, networks and build cache.
  4. Confirm which volumes contain application data before any volume cleanup.

Why this branch first: Docker provides docker system df specifically to report daemon disk use; host free-space/inode checks prevent misdiagnosing a non-Docker filesystem issue.

TroubleByte diagnostic diagram for Docker no space left on device, showing measurement, storage location and safe cleanup.
Docker storage-recovery pathTroubleByte · Original TroubleByte editorial diagram
Visual diagnosis

TroubleByte diagnostic path

01
Host

Check filesystem bytes and inodes.

02
Measure

Run docker system df.

03
Classify

Find reclaimable versus active data.

04
Prune

Remove only verified unused objects.

Original TroubleByte diagnostic map. It summarizes the cited troubleshooting order; it is not a vendor screenshot.

Decision checks

Use the symptom to choose the next branch

Does docker system df show substantial reclaimable unused data?

YES

Prune the specific unused category carefully.

NO

Investigate host filesystem, inodes, active logs/volumes or non-Docker usage.

Would the proposed cleanup include volumes?

YES

Verify each volume’s data ownership/backups before proceeding.

NO

Proceed with the narrower non-volume cleanup and remeasure.

Original TroubleByte decision aid derived from the cited troubleshooting scope. It does not replace vendor documentation.

Diagnosis

What this usually means

The error string does not prove Docker images are the only problem. First determine whether the host filesystem is full, inodes are exhausted, or Docker objects are the major consumer. Docker’s own df/prune tooling shows reclaimable data and avoids the unsafe pattern of manually deleting files under Docker’s data root.

What the evidence establishes

What we verified from the source material

Docker documents docker system df for measuring daemon disk usage and says unused objects are not removed automatically. Its pruning guidance explains exactly which object types are removed and that volumes are not pruned by default to protect data.

Before you change anything

Prerequisites and checks

Prepare first

  • Know whether the host runs production containers or important named volumes.
  • Have a current backup for data that cannot be recreated.
  • Do not manually delete files inside /var/lib/docker or the configured Docker data root.

Checks that prevent the wrong fix

  • Check host free space and inode availability on the filesystem containing Docker data.
  • Run docker system df (and -v if needed) to identify images, containers, volumes and cache usage.
  • Identify what is truly unused before any prune command.
Scope

Applies to

✓ Docker Engine✓ Linux hosts✓ Images and build cache✓ Containers and volumes
Troubleshooting path

Solutions, in order

01
SAFE · START HERE

Measure before deleting

  1. Check df -h and df -i for the Docker data filesystem.
  2. Run docker system df and optionally docker system df -v.
  3. Identify reclaimable images, stopped containers, networks and build cache.
  4. Confirm which volumes contain application data before any volume cleanup.

Why this can work: Docker provides docker system df specifically to report daemon disk use; host free-space/inode checks prevent misdiagnosing a non-Docker filesystem issue.

02
CAUTION

Prune only unused Docker objects

  1. Start with the narrowest prune command appropriate to the reclaimable category.
  2. Use docker system prune only after reading the confirmation list of objects it will remove.
  3. Do not add --volumes unless you have verified the affected unused volumes are safe to remove.
  4. Recheck docker system df and host free space after cleanup.

Why this can work: Docker intentionally uses conservative garbage collection and warns that system prune removes stopped containers, unused networks, dangling/unused images depending options, and build cache; volumes require explicit opt-in.

Verification

How to know the fix actually worked

  1. The host filesystem has adequate free bytes/inodes after cleanup.
  2. The failed Docker pull/build/container operation completes without the no-space error.

Do not count a temporary disappearance of the symptom as a confirmed fix if the problem normally returns after a restart, reconnect or several minutes of use.

Stop conditions

When not to keep changing things

  • If a volume or active writable layer contains important application data, stop broad prune/delete actions and investigate application retention/logging instead.

Escalation: If disk use is dominated by active container writable layers, logs or named volumes, stop broad pruning and investigate the specific application/data retention policy before deleting production data.

Evidence

Sources used for this record

Primary · checked Sep 8, 2026Docker — docker system dfFirst-party documentation used as the primary evidence source for this troubleshooting record.Corroborating · checked Sep 8, 2026Docker — Prune unused Docker objectsFirst-party documentation used as corroborating evidence for this troubleshooting record.
Who reviewed this

Mihailo Ivanjac

This record was written and source-reviewed by Mihailo Ivanjac. Source review means the cited documentation was checked against the troubleshooting order on this page; it does not imply a hands-on reproduction unless the page explicitly says so.

Author profile and editorial scope →
Community troubleshooting

Discuss this exact problem

Share what happened on your system, ask a focused follow-up question, or add evidence that may help someone with the same symptom. Community posts are separate from TroubleByte editorial verification.

Before posting Do not include passwords, API keys, recovery codes, private IP addresses or other secrets. TroubleByte automatically attaches only coarse OS, browser and device-class context. Your full Browser & System Info report is not uploaded automatically.
Start a discussion
Posts are public. Guest identity is pseudonymous and browser-specific.

Loading community discussions…
Keep diagnosing

Related Servers problems

DockerDocker container keeps restarting in a loopA Docker container that repeatedly returns to a running/restarting state can be caused by its main process exiting while a restart policy relaunches it. First inspect the container state, restart count and policy; also note that Docker only activates a restart policy after the container has started successfully for at least 10 seconds.DockerDocker container DNS resolution not workingDocker containers normally inherit or proxy DNS based on the network they use. Default-bridge containers receive resolver settings derived from the host, while custom networks use Docker’s embedded DNS. Loopback resolvers such as 127.0.0.1 on the host are a common mismatch because that address means the container itself inside its network namespace.ServersCannot connect to the Docker daemonDocker documents two primary causes for “Cannot connect to the Docker daemon”: the daemon is not running, or the client is pointed at another unreachable daemon. On Linux, local Unix-socket permissions can also prevent a non-root user from connecting even when dockerd is healthy.
Change log

Revision history

Show 1 recorded revision

2026-09-08 — Created for the v1.1 content-scale pass from current first-party documentation; public indexing is controlled by the launch allowlist.