a719472d94
Ironic is internally clustered and that causes some friction with Nova which assumes that each n-cpu instance is the only one responsible for any given VM. While this is going to be addressed in the long term, at the moment we need to make it possible to run in an HA setup for Ironic. The short term solution proposed is to run 2+ nova-compute's each of which reports the same hostname (e.g. ironic). This works but has some caveats - one of which is that _init_instance will now race between different nova-compute instances starting up at the same time. A custom ComputeManager permits us to address that without prejuidice to future long term solutions. Relatedly, TripleO needs Ironic to permit service startup before keystone is initialised, which the removal of API calls during init_host permits - and as there are no API calls needed for correct behaviour of the Ironic driver, this is straight forward :). See https://etherpad.openstack.org/p/ironic-nova-friction for more discussion. Change-Id: I68d46c4da8715df03c3a88393b55665dc57045a3 Closes-Bug: #1295503 |
||
|---|---|---|
| .. | ||
| apache2 | ||
| ironic | ||