1a57af20e8
By default we should prevent all neutron server going down at the same time Instead we should apply approach where we upgrade single neutron server and the rest of hosts next time to decrease API downtime Additionally changed default for neutron_other_serial, to safe 1 as in case of big amount of network nodes, several l3 agents may went down at the same time, and make ha router unavaliable. Change-Id: I618ea5083da29e7188c093c96f40ca3db6f9e5cc |
||
---|---|---|
.. | ||
cinder.yml | ||
glance.yml | ||
neutron.yml | ||
nova.yml |