Add ability for Sysbox to restart without affecting running containers

As of today, Sysbox requires user-intervention to restart system containers upon any of the Sysbox active components (sysbox-fs and sysbox-mgr) being restarted. At high-level, this requirement is a consequence of the direct association between the sys containers and the software constructs utilized to serve each container within the Sysbox daemons.

Ideally, no manual intervention should be required to achieve full service restoration, and more importantly, the container's disruption interval should be limited to the time it takes for Sysbox daemons to restart and complete their reconciliation logic.

As any high-availability feature, we expect this to be a relatively large effort that may potentially extend beyond a few release cycles.

Finally, due to the nature of this feature, we expect a considerable portion to be implemented as part of the Sysbox Enterprise product offering.

The issue does not name files, tests, or a specific entry point. Begin by tracing reconciliation behavior in sysbox-fs and sysbox-mgr, then determine the scope needed for daemon restarts without stopping running system containers. Done means service restoration occurs without manual container restarts and limits disruption to daemon recovery.