dc0 METAL-ADMIN DHCP HANDOVER -- Office1 region -> per-DC region (D-132 q1)
Captured 2026-07-30. Nodes were all Ready and POWERED OFF, so the gap was free.
======================================================================

-- region identity proven before the mutating step --
OK   profile 'vr1-dc0-region' -> region with racks [hot-kid] (as expected)

STEP 1  $ ssh voffice1 'maas admin vlan update 4 0 dhcp_on=false'
  read back:
    OFFICE1 10.12.8.0/22 dhcp_on=False primary_rack=7chphy fabric=fabric-4

STEP 2  verify STOPPED by PROCESS (MAAS self-report is not evidence --
        the 2026-07-20 Temporal incident class)
    dc0 rack pgrep -c dhcpd = 0

STEP 3  $ ssh voffice1 'maas vr1-dc0-region vlan update 0 0 dhcp_on=true primary_rack=c3aqh8'
    NEW REGION 10.12.8.0/22 dhcp_on=True primary_rack=c3aqh8 fabric=fabric-0 space=metal-admin

STEP 4  verify RUNNING by PROCESS, and that the rack did NOT come back
    region VM (hot-kid):
      root       20565    5585  0 21:28 ?        00:00:00 /snap/maas/41649/usr/sbin/dhcpd -f -4 -pf /var/snap/maas/common/maas/dhcp/dhcpd.pid -cf /var/snap/maas/common/maas/dhcpd.conf -lf /var/snap/maas/common/maas/dhcp/dhcpd.leases enp1s0
      root       20571    5585  0 21:28 ?        00:00:00 /snap/maas/41649/usr/sbin/dhcpd -f -6 -pf /var/snap/maas/common/maas/dhcp/dhcpd6.pid -cf /var/snap/maas/common/maas/dhcpd6.conf -lf /var/snap/maas/common/maas/dhcp/dhcpd6.leases enp1s0
    dc0 rack pgrep -c dhcpd = 0

RESULT: exactly ONE DHCP server on the dc0 metal-admin segment, owned by the
DC-LOCAL region. dhcpd6 also starts (the old region ran v4 only) but the v6
metal-admin subnet carries NO dynamic range -- only rfc-4291-2.6.1 and
assigned-ip/reserved -- so it has nothing to hand out and nodes keep their
ruled MAAS statics. Recorded so a later session does not read it as a fault.