Bad exit status for docker system info on failure

[fornellas-maze]

Description

docker system info returns 0 (success) even on failures.

Reproduce

$ DOCKER_HOST=unix:///non-existent docker system info --format '{{.Architecture}}' ; echo $?
Cannot connect to the Docker daemon at unix:///non-existent. Is the docker daemon running?
0

Expected behavior

on failures, any exit value other than 0 would be appropriate.

docker version

Client: Docker Engine - Community
 Version:           28.0.1
 API version:       1.48
 Go version:        go1.23.6
 Git commit:        068a01e
 Built:             Wed Feb 26 10:41:12 2025
 OS/Arch:           linux/amd64
 Context:           default

Server: Docker Engine - Community
 Engine:
  Version:          28.0.1
  API version:      1.48 (minimum version 1.24)
  Go version:       go1.23.6
  Git commit:       bbd0a17
  Built:            Wed Feb 26 10:41:12 2025
  OS/Arch:          linux/amd64
  Experimental:     false
 containerd:
  Version:          1.7.25
  GitCommit:        bcc810d6b9066471b0b6fa75f557a15a1cbf31bb
 runc:
  Version:          1.2.4
  GitCommit:        v1.2.4-0-g6c52b3f
 docker-init:
  Version:          0.19.0
  GitCommit:        de40ad0

docker info

Client: Docker Engine - Community
 Version:    28.0.1
 Context:    default
 Debug Mode: false
 Plugins:
  buildx: Docker Buildx (Docker Inc.)
    Version:  v0.21.1
    Path:     /usr/libexec/docker/cli-plugins/docker-buildx
  compose: Docker Compose (Docker Inc.)
    Version:  v2.33.1
    Path:     /usr/libexec/docker/cli-plugins/docker-compose

Server:
 Containers: 1
  Running: 1
  Paused: 0
  Stopped: 0
 Images: 29
 Server Version: 28.0.1
 Storage Driver: overlay2
  Backing Filesystem: extfs
  Supports d_type: true
  Using metacopy: false
  Native Overlay Diff: true
  userxattr: false
 Logging Driver: json-file
 Cgroup Driver: systemd
 Cgroup Version: 2
 Plugins:
  Volume: local
  Network: bridge host ipvlan macvlan null overlay
  Log: awslogs fluentd gcplogs gelf journald json-file local splunk syslog
 Swarm: inactive
 Runtimes: io.containerd.runc.v2 runc
 Default Runtime: runc
 Init Binary: docker-init
 containerd version: bcc810d6b9066471b0b6fa75f557a15a1cbf31bb
 runc version: v1.2.4-0-g6c52b3f
 init version: de40ad0
 Security Options:
  apparmor
  seccomp
   Profile: builtin
  cgroupns
 Kernel Version: 6.8.0-55-generic
 Operating System: KDE neon 6.3
 OSType: linux
 Architecture: x86_64
 CPUs: 16
 Total Memory: 30.66GiB
 Name: framework
 ID: ed7176be-9632-4d36-b469-32b4cbaab421
 Docker Root Dir: /var/lib/docker
 Debug Mode: false
 Experimental: false
 Insecure Registries:
  ::1/128
  127.0.0.0/8
 Live Restore Enabled: false

Additional Info

I'm attempting to use this to manage a dev container, and define the architecture, but when it fails... it is hard to know, and unexpected behaviour happens down the line.

[thaJeztah]

Thanks for reporting; this looks be related to the --format template used; running the same command, but without a format set does return a non-zero exit code;

DOCKER_HOST=unix:///non-existent docker system info --format '{{.Architecture}}'
echo $?
0

DOCKER_HOST=unix:///non-existent docker system info
# (output omitted)

echo $?
1

I know there's some code in this area that tries to parse the template to detect whether a daemon connection is required (so that client-side info could still be printed, even with the daemon down). It wouldn't surprise me if something in that area of the code is swallowing the exit code 🤔

[thaJeztah]

If you only need the architecture (and/or OS), I would recommend using docker version; it's a much more lightweight endpoint (docker info does a lot behind the scenes to collect info, including traversing directories to collect version information for CLI plugins);

docker version --format='{{.Server.Arch}}'
arm64

docker version --format='{{.Server.Os}}'
linux

The architecture returned there is using the formats used by Golang's GOOS, which also matches the format used for platform strings in OCI images.

Comparison in time taken for both;

$ time docker version --format='{{.Server.Arch}}'
arm64

real	0m0.040s
user	0m0.011s
sys	0m0.008s

$ time docker info --format='{{.Architecture}}'
aarch64

real	0m0.272s
user	0m0.158s
sys	0m0.391s

[Benehiko]

I see what the issue is here, the info command prints the error itself and does not propagate the error up the stack to the main function. I'll see if I can do a quick fix for this.
https://github.com/docker/cli/blob/master/cli/command/system/info.go#L113

[thaJeztah]

Ah! Yes, looks indeed in one of the areas I had in mind. It's a bit gnarly for sure; I think it was in an attempt to make the info do as much as possible, even if there's some failure (as the information needed may be client-only). I guess we need to store the error and if no other error is returned, return it (and/or an errors.Join for multiple errors 🤔)