fix: the clone identity reset could never reboot, so it never took effect #30
No reviewers
Labels
No labels
blocked
blocker:ci-red
blocker:conflict
blocker:drill-pending
blocker:unrequested
bug
claimed
documentation
enhancement
epic
merge-next
needs-triage
ready
release
scope:cli
scope:drill
scope:host
scope:installer
scope:templates
scope:tiers
stale
state:addressing
state:bots-reviewing
state:building
state:needs-human
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference: heavy-duty/box#30
Loading…
Reference in a new issue
No description provided.
Delete branch "fix/reset-identity-restart"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
My fix in #27 was right about the cause and wrong about the mechanism. Run 7:
systemdneeds a valid/etc/machine-idto shut down cleanly. #27 truncated it and then calledincus restart— so the graceful stop hung, Incus timed out, and the reboot never happened. The new machine-id only takes effect at boot, so the clone kept its source's identity: the exact bug the reset exists to prevent.And it cascaded
The duplicate address didn't just break cloning — it broke the box. Run 7's audit shows
box cannot reach the internet(two boxes fighting over10.87.0.215), and phase D then reportedmac+ipv4 filtering BROKE the box's networking — design vetoagainst #16. That verdict is false; the box was already broken. (Guarded separately in the drill.)The fix
Write a fresh valid id rather than emptying the file:
systemd-machine-id-setupderives it from the DMI product UUID, which Incus makes unique per instance — so the clone gets an identity that is genuinely its own, not a random one. Then restart with a real timeout and a forced fallback:A clone that keeps its source's DHCP lease is worse than an unclean stop of a box that booted thirty seconds ago.
Verify
🤖 Generated with Claude Code