Repository navigation
Commit 48471ad
[docs] blue/green: stop all OpenStack services on A before the backup
After the drain, all OpenStack services on A are scaled to zero while
A's infrastructure (Galera, RabbitMQ, memcached, OVN, dnsmasq) keeps
running. The backup becomes a real freeze with a consistent cut across
all Galera clusters, and only B can act on shared backends afterwards.
This removes the duplicate-actor risks from the residual risk list and
shrinks B's fence list to the reactions to EDPM not being attached yet:
OVN gateway, InstanceHA, octavia health-manager/housekeeping, aodh,
heat-engine and watcher. EDPM stays attached to A's infrastructure until
the switch, so north-south traffic through ctlplane gateways keeps
working during the window.
Rollback before the switch only restarts A's services. After the switch
it follows the switch sequence with roles swapped. A's DBs are unchanged
since the stop.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>1 parent 8ca880f commit 48471ad
1 file changed
Lines changed: 46 additions & 34 deletions
0 commit comments