[bug]: space-commercial v2.6.1 container healthcheck fails because curl is missing

Is there an existing issue for this?

• I have searched the existing issues

Related closed issues/PRs:

• #8392
• #8671
• #8674

Those appear to address this for the public Space Dockerfile, but the commercial image still reproduces the missing-curl failure.

Current behavior

On a self-hosted commercial Plane deployment upgraded to v2.6.1, the plane-space-1 container remains unhealthy even though the Space app responds successfully on /spaces/.

The image is:

makeplane/space-commercial:v2.6.1

The container healthcheck is:

curl -fsS http://127.0.0.1:3000/spaces/ >/dev/null || exit 1

But curl is not installed inside the container, so Docker marks the container unhealthy.

Health log output:

exit=1 output="/bin/sh: curl: not found\n"

The service itself appears responsive. These checks succeed inside the same container:

wget -q -O /dev/null http://127.0.0.1:3000/spaces/
node -e "fetch('http://127.0.0.1:3000/spaces/').then(r=>process.exit(r.ok?0:1)).catch(()=>process.exit(1))"

Expected behavior

makeplane/space-commercial:v2.6.1 should report healthy when the Space app is responding.

Either:

1. include curl in the commercial Space image, matching the public Dockerfile fix from #8674, or
2. change the healthcheck to use a runtime dependency already guaranteed in the Space image, such as Node.

Steps to reproduce

1. Run a Docker Compose self-hosted commercial deployment using makeplane/space-commercial:v2.6.1.
2. Wait for containers to start.
3. Check container status:

docker ps --filter name=plane-space

4. Inspect health logs:

docker inspect --format '{{range .State.Health.Log}}{{.End}} exit={{.ExitCode}} output={{printf "%q" .Output}}{{"\n"}}{{end}}' plane-space-1

Environment

• Variant: Self-hosted commercial
• Deployment: Docker Compose
• Plane version: v2.6.1
• Space image: makeplane/space-commercial:v2.6.1
• Host OS: Ubuntu 24.04 LTS

Impact

The Space service can be functional, but Docker reports it unhealthy. This creates false alarms and can mislead upgrade/runbook checks or orchestrators that rely on container health state.

Suggested fix

Ensure the commercial Space image includes the same healthcheck fix as the public Dockerfile, or use a healthcheck that does not depend on curl.

For example, a Node-based healthcheck works in the current image:

node -e "fetch('http://127.0.0.1:3000/spaces/').then(r=>process.exit(r.ok?0:1)).catch(()=>process.exit(1))"