Lesson 12 · Docker CLI

The Docker CLI day to day: inspect, filter, format and clean up

The commands you use every day for images, containers, volumes and networks, how to filter and format their output for scripts, and how to win back disk space safely with docker system df and prune.

25 min read · beginner · hands-on

By the end of this lesson you'll be able to

0 of 5 complete

Manage images, containers, volumes and networks with the docker <object> <verb> commands
Filter lists with --filter and shape output with --format and Go templates
Read any setting of a container or image with docker inspect
Copy files in and out of containers, and watch live usage with docker stats
See where disk space goes with docker system df, and reclaim it with the right prune

One pattern

Every Docker command is docker <object> <verb>. Learn the verbs once and they work everywhere:

Object	ls	inspect	rm	prune
`image`	`docker image ls` (`docker images`)	`docker image inspect`	`docker image rm` (`rmi`)	`docker image prune`
`container`	`docker container ls` (`docker ps`)	`docker container inspect`	`docker container rm` (`rm`)	`docker container prune`
`volume`	`docker volume ls`	`docker volume inspect`	`docker volume rm`	`docker volume prune`
`network`	`docker network ls`	`docker network inspect`	`docker network rm`	`docker network prune`

docker <object> --help lists the rest.

Images

$ docker image ls --filter reference='price-api*' --format '{{.Repository}}:{{.Tag}} {{.Size}}'
price-api:multi 14.6MB
price-api:dev 14.6MB
price-api:single 508MB
$ docker image history price-api:multi            # layers and their sizes (lesson 7)
$ docker image inspect price-api:multi --format '{{.Config.User}} {{json .Config.Entrypoint}}'
nonroot ["/price-api"]
$ docker image rm price-api:single                # removes the tag; layers go when nothing uses them
$ docker pull --platform linux/arm64 alpine:3.22  # a specific architecture

docker image ls --filter reference='price-api*' --format '{{.Repository}}:{{.Tag}} {{.Size}}'

dockerbuild and run containers
imagemanage images
lslist them
--filter reference='price-api*'show only matching items
--format '{{.Repository}}:{{.Tag}} {{.Size}}'print only these fields (Go template)

docker image inspect price-api:multi --format '{{.Config.User}} {{json .Config.Entrypoint}}'

inspectshow details as JSON
--format '{{.Config.User}} {{json .Config.Entrypoint}}'print only these fields (Go template)

docker image rm price-api:single

rmdelete one

docker pull --platform linux/arm64 alpine:3.22

pulldownload an image
--platform linux/arm64which CPU architecture
alpine:3.22the image

Containers

$ docker ps -a --filter status=exited --format 'table {{.Names}}\t{{.Status}}'
$ docker inspect web --format '{{range .NetworkSettings.Networks}}{{.IPAddress}}{{end}}'
$ docker logs --since 10m --tail 50 web
$ docker exec -it web sh
$ docker cp web:/etc/nginx/nginx.conf ./nginx.conf       # copy out of a container
$ docker cp ./index.html web:/usr/share/nginx/html/      # and into it
$ docker stats --no-stream                               # CPU, memory, network, against limits
$ docker top web                                         # processes, as the host sees them
$ docker container rm -f $(docker ps -aq --filter label=env=test)   # remove a whole group

docker ps -a --filter status=exited --format 'table {{.Names}}\t{{.Status}}'

dockerbuild and run containers
pslist running containers
-ainclude stopped ones too
--filter status=exitedshow only matching items
--format 'table {{.Names}}\t{{.Status}}'print only these fields (Go template)

docker inspect web --format '{{range .NetworkSettings.Networks}}{{.IPAddress}}{{end}}'

inspectshow every detail as JSON
webthe container or image
--format '{{range .NetworkSettings.Networks}}{{.IPAddress}}{{end}}'print only these fields (Go template)

docker logs --since 10m --tail 50 web

logsshow a container’s output
--since 10monly output newer than this
--tail 50only the last N lines
webthe container

docker exec -it web sh

execrun a command in a running container
-itinteractive, with a terminal
shthe command to run inside it

docker cp web:/etc/nginx/nginx.conf ./nginx.conf

cpcopy files between a container and the host

docker stats --no-stream

statslive CPU and memory use
--no-streamprint once instead of updating

docker top web

topprocesses in a container

docker container rm -f $(docker ps -aq --filter label=env=test)

containermanage containers
rmdelete one
-fdon’t ask to confirm
$(docker ps -aq --filter label=env=test)the output of another command

--filter narrows lists (by status, name, label, ancestor, network…). --format takes a Go template: {{.Names}}, {{json .Config}}, table for aligned columns. docker inspect returns everything Docker knows about an object as JSON, and is the first place to look when something behaves oddly: the mounts, environment, restart count, health log and exit code are all there.

Volumes and networks

$ docker volume ls --filter dangling=true          # volumes no container uses
$ docker volume inspect shop_pgdata --format '{{.Mountpoint}}'
$ docker network ls
$ docker network inspect appnet --format '{{range .Containers}}{{.Name}} {{.IPv4Address}}{{"\n"}}{{end}}'

docker volume ls --filter dangling=true

dockerbuild and run containers
volumemanage volumes
lslist them
--filter dangling=trueshow only matching items

docker volume inspect shop_pgdata --format '{{.Mountpoint}}'

inspectshow details as JSON
--format '{{.Mountpoint}}'print only these fields (Go template)

docker network ls

networkmanage networks

docker network inspect appnet --format '{{range .Containers}}{{.Name}} {{.IPv4Address}}{{"\n"}}{{end}}'

--format '{{range .Containers}}{{.Name}} {{.IPv4Address}}{{"\n"}}{{end}}'print only these fields (Go template)

Cleanup

Docker never deletes anything on its own. Images from old builds, stopped containers, anonymous volumes and the build cache pile up until the disk is full, often on a CI runner at 2 a.m. See where the space goes:

$ docker system df
TYPE            TOTAL     ACTIVE    SIZE      RECLAIMABLE
Images          21        0         5.446GB   479MB (8%)
Containers      0         0         0B        0B
Local Volumes   14        0         510.6MB   510.6MB (100%)
Build Cache     74        0         1.226GB   447.7MB

docker system df

dockerbuild and run containers
systemdisk use and cleanup
dfdisk use by type

That's this machine after the course so far: 14 volumes nobody uses (mostly anonymous Postgres volumes from lesson 6) and half a gigabyte of build cache. Clean each kind on purpose:

$ docker volume prune -f
Total reclaimed space: 510.6MB
$ docker builder prune -f
Total:	404.6MB
$ docker system df
TYPE            TOTAL     ACTIVE    SIZE      RECLAIMABLE
Images          21        0         5.045GB   77.44MB (1%)
Containers      0         0         0B        0B
Local Volumes   1         0         0B        0B
Build Cache     37        0         820.9MB   43.11MB

docker volume prune -f

dockerbuild and run containers
volumemanage volumes
prunedelete the unused ones
-fdon’t ask to confirm

docker builder prune -f

buildermanage the build cache

docker system df

systemdisk use and cleanup
dfdisk use by type

Command	Removes
`docker container prune`	stopped containers
`docker image prune`	dangling images (untagged leftovers); `-a`: every image no container uses
`docker volume prune`	unused anonymous volumes; `--all`: unused named volumes too
`docker network prune`	networks no container uses
`docker builder prune`	build cache; `--filter until=168h` keeps the last week
`docker system prune`	containers, networks, dangling images and build cache; add `--volumes` or `-a` to go further

Measure before you prune

docker system df maps image, container and volume usage; prune narrowly after checking what is reclaimable. inspect --format gives stable fields for scripts, and docker context show confirms which daemon a command will affect. These are operational safeguards, not backups.

$ docker system df
$ docker ps --format '{{.Names}} {{.Image}} {{.Status}}'
$ docker inspect api --format '{{.Name}} {{range .NetworkSettings.Networks}}{{.IPAddress}}{{end}}'
$ docker context show
$ docker container prune --filter until=24h

docker system df

dockerbuild and run containers
systemdisk use and cleanup
dfdisk use by type

docker ps --format '{{.Names}} {{.Image}} {{.Status}}'

pslist running containers
--format '{{.Names}} {{.Image}} {{.Status}}'print only these fields (Go template)

docker inspect api --format '{{.Name}} {{range .NetworkSettings.Networks}}{{.IPAddress}}{{end}}'

inspectshow every detail as JSON
apithe container or image
--format '{{.Name}} {{range .NetworkSettings.Networks}}{{.IPAddress}}{{end}}'print only these fields (Go template)

docker context show

contextchoose which daemon to talk to

docker container prune --filter until=24h

containermanage containers
prunedelete the unused ones
--filter until=24hshow only matching items

Mini mission

Find and reclaim wasted space

Measure Docker's disk use, create some junk on purpose (exited containers, anonymous volumes, a dangling image), find each kind with filters, and remove it with the matching prune without touching a named volume you want to keep.

Input: DockerOutput: docker system df before and after, with the keep-me volume still present

Reveal one safe solution

$ docker system df
$ docker volume create keep-me
$ for i in 1 2 3; do docker run -v /data alpine:3.22 true; done        # 3 exited containers, 3 anonymous volumes
$ docker ps -a --filter status=exited --filter ancestor=alpine:3.22 --format '{{.Names}} {{.Status}}'
$ docker container prune -f
$ docker volume ls --filter dangling=true        # now unused: the 3 anonymous ones, and keep-me
$ docker volume prune -f
$ docker volume ls --filter name=keep-me
DRIVER    VOLUME NAME
local     keep-me
$ docker system df && docker volume rm keep-me

docker system df

dockerbuild and run containers
systemdisk use and cleanup
dfdisk use by type

docker volume create keep-me

volumemanage volumes
createcreate one

for i in 1 2 3; do docker run -v /data alpine:3.22 true; done

forloop over a list
dostart of what the loop runs each time
runstart a new container
-v /datamount a volume or folder, source:path-inside
alpine:3.22the image to start
true;the command run inside, instead of the image’s default
doneend of the loop

docker ps -a --filter status=exited --filter ancestor=alpine:3.22 --format '{{.Names}} {{.Status}}'

pslist running containers
-ainclude stopped ones too
--filter status=exitedshow only matching items
--filter ancestor=alpine:3.22show only matching items
--format '{{.Names}} {{.Status}}'print only these fields (Go template)

docker container prune -f

containermanage containers
prunedelete the unused ones
-fdon’t ask to confirm

docker volume ls --filter dangling=true

lslist them
--filter dangling=trueshow only matching items

docker volume ls --filter name=keep-me

--filter name=keep-meshow only matching items

docker system df && docker volume rm keep-me

&&run the next command only if this one worked
rmdelete one

volume prune removed the anonymous volumes but kept keep-me, because named volumes need --all. Order matters too: volumes still attached to a stopped container aren't unused, so containers go first.

Try it yourself

0 of 6 steps done

Run each task in your own account or terminal, then tick it off.

Run docker system df and find which type uses the most reclaimable space.
List only exited containers with docker ps -a --filter status=exited.
Print each container's name and IP address with a --format template.
Copy a file out of a running nginx container with docker cp.
Watch docker stats --no-stream while a container is busy.
Run docker volume prune and read what it removed before confirming.

Knowledge check

Quick check

Question 1 of 5

docker system df shows Local Volumes at 510MB, 100% reclaimable. What does that mean?

Go deeper

Official reading for this lesson

Or get every quiz answer right and it completes itself.

Lesson 12 · Docker CLI

The Docker CLI day to day: inspect, filter, format and clean up

The commands you use every day for images, containers, volumes and networks, how to filter and format their output for scripts, and how to win back disk space safely with docker system df and prune.

25 min read · beginner · hands-on

By the end of this lesson you'll be able to

0 of 5 complete

Manage images, containers, volumes and networks with the docker <object> <verb> commands
Filter lists with --filter and shape output with --format and Go templates
Read any setting of a container or image with docker inspect
Copy files in and out of containers, and watch live usage with docker stats
See where disk space goes with docker system df, and reclaim it with the right prune

One pattern

Every Docker command is docker <object> <verb>. Learn the verbs once and they work everywhere:

Object	ls	inspect	rm	prune
`image`	`docker image ls` (`docker images`)	`docker image inspect`	`docker image rm` (`rmi`)	`docker image prune`
`container`	`docker container ls` (`docker ps`)	`docker container inspect`	`docker container rm` (`rm`)	`docker container prune`
`volume`	`docker volume ls`	`docker volume inspect`	`docker volume rm`	`docker volume prune`
`network`	`docker network ls`	`docker network inspect`	`docker network rm`	`docker network prune`

docker <object> --help lists the rest.

Images

$ docker image ls --filter reference='price-api*' --format '{{.Repository}}:{{.Tag}} {{.Size}}'
price-api:multi 14.6MB
price-api:dev 14.6MB
price-api:single 508MB
$ docker image history price-api:multi            # layers and their sizes (lesson 7)
$ docker image inspect price-api:multi --format '{{.Config.User}} {{json .Config.Entrypoint}}'
nonroot ["/price-api"]
$ docker image rm price-api:single                # removes the tag; layers go when nothing uses them
$ docker pull --platform linux/arm64 alpine:3.22  # a specific architecture

docker image ls --filter reference='price-api*' --format '{{.Repository}}:{{.Tag}} {{.Size}}'

dockerbuild and run containers
imagemanage images
lslist them
--filter reference='price-api*'show only matching items
--format '{{.Repository}}:{{.Tag}} {{.Size}}'print only these fields (Go template)

docker image inspect price-api:multi --format '{{.Config.User}} {{json .Config.Entrypoint}}'

inspectshow details as JSON
--format '{{.Config.User}} {{json .Config.Entrypoint}}'print only these fields (Go template)

docker image rm price-api:single

rmdelete one

docker pull --platform linux/arm64 alpine:3.22

pulldownload an image
--platform linux/arm64which CPU architecture
alpine:3.22the image

Containers

$ docker ps -a --filter status=exited --format 'table {{.Names}}\t{{.Status}}'
$ docker inspect web --format '{{range .NetworkSettings.Networks}}{{.IPAddress}}{{end}}'
$ docker logs --since 10m --tail 50 web
$ docker exec -it web sh
$ docker cp web:/etc/nginx/nginx.conf ./nginx.conf       # copy out of a container
$ docker cp ./index.html web:/usr/share/nginx/html/      # and into it
$ docker stats --no-stream                               # CPU, memory, network, against limits
$ docker top web                                         # processes, as the host sees them
$ docker container rm -f $(docker ps -aq --filter label=env=test)   # remove a whole group

docker ps -a --filter status=exited --format 'table {{.Names}}\t{{.Status}}'

dockerbuild and run containers
pslist running containers
-ainclude stopped ones too
--filter status=exitedshow only matching items
--format 'table {{.Names}}\t{{.Status}}'print only these fields (Go template)

docker inspect web --format '{{range .NetworkSettings.Networks}}{{.IPAddress}}{{end}}'

inspectshow every detail as JSON
webthe container or image
--format '{{range .NetworkSettings.Networks}}{{.IPAddress}}{{end}}'print only these fields (Go template)

docker logs --since 10m --tail 50 web

logsshow a container’s output
--since 10monly output newer than this
--tail 50only the last N lines
webthe container

docker exec -it web sh

execrun a command in a running container
-itinteractive, with a terminal
shthe command to run inside it

docker cp web:/etc/nginx/nginx.conf ./nginx.conf

cpcopy files between a container and the host

docker stats --no-stream

statslive CPU and memory use
--no-streamprint once instead of updating

docker top web

topprocesses in a container

docker container rm -f $(docker ps -aq --filter label=env=test)

containermanage containers
rmdelete one
-fdon’t ask to confirm
$(docker ps -aq --filter label=env=test)the output of another command

Volumes and networks

$ docker volume ls --filter dangling=true          # volumes no container uses
$ docker volume inspect shop_pgdata --format '{{.Mountpoint}}'
$ docker network ls
$ docker network inspect appnet --format '{{range .Containers}}{{.Name}} {{.IPv4Address}}{{"\n"}}{{end}}'

docker volume ls --filter dangling=true

dockerbuild and run containers
volumemanage volumes
lslist them
--filter dangling=trueshow only matching items

docker volume inspect shop_pgdata --format '{{.Mountpoint}}'

inspectshow details as JSON
--format '{{.Mountpoint}}'print only these fields (Go template)

docker network ls

networkmanage networks

docker network inspect appnet --format '{{range .Containers}}{{.Name}} {{.IPv4Address}}{{"\n"}}{{end}}'

--format '{{range .Containers}}{{.Name}} {{.IPv4Address}}{{"\n"}}{{end}}'print only these fields (Go template)

Cleanup

$ docker system df
TYPE            TOTAL     ACTIVE    SIZE      RECLAIMABLE
Images          21        0         5.446GB   479MB (8%)
Containers      0         0         0B        0B
Local Volumes   14        0         510.6MB   510.6MB (100%)
Build Cache     74        0         1.226GB   447.7MB

docker system df

dockerbuild and run containers
systemdisk use and cleanup
dfdisk use by type

That's this machine after the course so far: 14 volumes nobody uses (mostly anonymous Postgres volumes from lesson 6) and half a gigabyte of build cache. Clean each kind on purpose:

$ docker volume prune -f
Total reclaimed space: 510.6MB
$ docker builder prune -f
Total:	404.6MB
$ docker system df
TYPE            TOTAL     ACTIVE    SIZE      RECLAIMABLE
Images          21        0         5.045GB   77.44MB (1%)
Containers      0         0         0B        0B
Local Volumes   1         0         0B        0B
Build Cache     37        0         820.9MB   43.11MB

docker volume prune -f

dockerbuild and run containers
volumemanage volumes
prunedelete the unused ones
-fdon’t ask to confirm

docker builder prune -f

buildermanage the build cache

docker system df

systemdisk use and cleanup
dfdisk use by type

Command	Removes
`docker container prune`	stopped containers
`docker image prune`	dangling images (untagged leftovers); `-a`: every image no container uses
`docker volume prune`	unused anonymous volumes; `--all`: unused named volumes too
`docker network prune`	networks no container uses
`docker builder prune`	build cache; `--filter until=168h` keeps the last week
`docker system prune`	containers, networks, dangling images and build cache; add `--volumes` or `-a` to go further

Measure before you prune

$ docker system df
$ docker ps --format '{{.Names}} {{.Image}} {{.Status}}'
$ docker inspect api --format '{{.Name}} {{range .NetworkSettings.Networks}}{{.IPAddress}}{{end}}'
$ docker context show
$ docker container prune --filter until=24h

docker system df

dockerbuild and run containers
systemdisk use and cleanup
dfdisk use by type

docker ps --format '{{.Names}} {{.Image}} {{.Status}}'

pslist running containers
--format '{{.Names}} {{.Image}} {{.Status}}'print only these fields (Go template)

docker inspect api --format '{{.Name}} {{range .NetworkSettings.Networks}}{{.IPAddress}}{{end}}'

inspectshow every detail as JSON
apithe container or image
--format '{{.Name}} {{range .NetworkSettings.Networks}}{{.IPAddress}}{{end}}'print only these fields (Go template)

docker context show

contextchoose which daemon to talk to

docker container prune --filter until=24h

containermanage containers
prunedelete the unused ones
--filter until=24hshow only matching items

Mini mission

Find and reclaim wasted space

Input: DockerOutput: docker system df before and after, with the keep-me volume still present

Reveal one safe solution

$ docker system df
$ docker volume create keep-me
$ for i in 1 2 3; do docker run -v /data alpine:3.22 true; done        # 3 exited containers, 3 anonymous volumes
$ docker ps -a --filter status=exited --filter ancestor=alpine:3.22 --format '{{.Names}} {{.Status}}'
$ docker container prune -f
$ docker volume ls --filter dangling=true        # now unused: the 3 anonymous ones, and keep-me
$ docker volume prune -f
$ docker volume ls --filter name=keep-me
DRIVER    VOLUME NAME
local     keep-me
$ docker system df && docker volume rm keep-me

docker system df

dockerbuild and run containers
systemdisk use and cleanup
dfdisk use by type

docker volume create keep-me

volumemanage volumes
createcreate one

for i in 1 2 3; do docker run -v /data alpine:3.22 true; done

forloop over a list
dostart of what the loop runs each time
runstart a new container
-v /datamount a volume or folder, source:path-inside
alpine:3.22the image to start
true;the command run inside, instead of the image’s default
doneend of the loop

docker ps -a --filter status=exited --filter ancestor=alpine:3.22 --format '{{.Names}} {{.Status}}'

pslist running containers
-ainclude stopped ones too
--filter status=exitedshow only matching items
--filter ancestor=alpine:3.22show only matching items
--format '{{.Names}} {{.Status}}'print only these fields (Go template)

docker container prune -f

containermanage containers
prunedelete the unused ones
-fdon’t ask to confirm

docker volume ls --filter dangling=true

lslist them
--filter dangling=trueshow only matching items

docker volume ls --filter name=keep-me

--filter name=keep-meshow only matching items

docker system df && docker volume rm keep-me

&&run the next command only if this one worked
rmdelete one

Try it yourself

0 of 6 steps done

Run each task in your own account or terminal, then tick it off.

Run docker system df and find which type uses the most reclaimable space.
List only exited containers with docker ps -a --filter status=exited.
Print each container's name and IP address with a --format template.
Copy a file out of a running nginx container with docker cp.
Watch docker stats --no-stream while a container is busy.
Run docker volume prune and read what it removed before confirming.

Knowledge check

Quick check

Question 1 of 5

docker system df shows Local Volumes at 510MB, 100% reclaimable. What does that mean?

Go deeper

Official reading for this lesson

Or get every quiz answer right and it completes itself.