Bad wording/missing information for live restore and swarm mode #26424

Is this a docs issue?

• My issue is about the documentation content or website

Type of issue

Information is incorrect

Description

Hi there,
your original text for having live restore and swarm mode is incorrect or better said, inconsistent.

The way the paragraph is written suggests that in mixed-case scenario (which is perfectly legal), containers running as standalone will be managed with live restore, while containers running as services in swarm mode, would be managed by swarm.

Therefore, if one would enable live restore in daemon.js on a system that is both a swarm member and has standalone containers, one would expect both to work.

That is not true because docker implementation straight-up refuses to start the daemon if both live restore are enabled and node is part of a swarm:

Oct 23 12:28:39 msi dockerd[3191]: time="2026-10-23T12:28:39.195135279+02:00" level=info msg="Loading containers: done." Oct 23 12:28:39 msi dockerd[3191]: time="2026-10-23T12:28:39.218291222+02:00" level=info msg="Docker daemon" commit=0367cbe containerd-snapshotter=false storage-driver=overlay2 version=29.9.2 Oct 23 12:28:39 msi dockerd[3191]: time="2026-10-23T12:28:39.218905151+02:00" level=info msg="Daemon shutdown complete" error="failed to start cluster component: --live-restore daemon configuration is incompatible with swarm mode" Oct 23 12:28:39 msi dockerd[3191]: time="2026-10-23T12:28:39.218938380+02:00" level=info msg="stopping event stream following graceful shutdown" error="context canceled" module=libcontainerd namespace=moby Oct 23 12:28:39 msi dockerd[3191]: failed to start cluster component: --live-restore daemon configuration is incompatible with swarm mode Oct 23 12:28:39 msi dockerd[3191]: time="2026-10-23T12:28:39.218939560+02:00" level=info msg="stopping event stream following graceful shutdown" error="context canceled" module=libcontainerd namespace=plugins.moby Oct 23 12:28:39 msi systemd[1]: docker.service: Main process exited, code=exited, status=1/FAILURE

Location

https://docs.docker.com/engine/daemon/live-restore/

Suggestion

Please rewrite the live restore and swarm paragraph to explicitly say that trying to enable live restore on a node that is already a swarm member will break running docker daemon, and that it has to be only one of them: live restore is only for non-swarm node (no mixed mode possible) or if node is swarm member, live restore is not possible.
Or something similar but very precise what is allowed and what will break your system.

Of course, the best solution would be to have both options working for mixed mode cases, where docker would simply ignore swarm orchestrated containers on a swarm node and just manage standalone containers when live restore is enabled.