In our Prod environment we currently use two Cinder Volume AZs.
The first AZ was setup via SOC Crowbar.
For the second AZ we used the stacked configuration file future and replaced some cinder.conf parameters.
But therefor we still use the non-clustered (active-passive) Cinder-Volume configuration where the Cinder-Volume services are pinned to a specific control node:
root@d2c-76-8a-ac-64-6e:~ # cinder-manage service list
Binary Host Zone Status State Updated At
cinder-scheduler d2c-76-8a-ac-65-ff nova enabled :-) 2017-03-28 10:04:38.919489
cinder-volume d2c-76-8a-ac-64-ff@backend-netapp-lab42bAZ2-0 lab42bAZ2 enabled :-) 2017-03-28 10:04:39.045227
cinder-scheduler d2c-76-8a-ac-64-ff nova enabled :-) 2017-03-28 10:04:42.590909
cinder-volume d2c-76-8a-ac-64-ff@backend-netapp-0 nova enabled :-) 2017-03-28 10:04:43.464835
cinder-scheduler d2c-76-8a-ac-64-ff nova enabled :-) 2017-03-28 10:04:39.567708
root@d2c-76-8a-ac-64-ff:~ #
Can you please add support of multiple Cinder-Volume services in different Cinder-AZs in the same control cluster with active-passive failover support.
In our Prod environment we currently use two Cinder Volume AZs. The first AZ was setup via SOC Crowbar. For the second AZ we used the stacked configuration file future and replaced some cinder.conf parameters.
But therefor we still use the non-clustered (active-passive) Cinder-Volume configuration where the Cinder-Volume services are pinned to a specific control node:
Can you please add support of multiple Cinder-Volume services in different Cinder-AZs in the same control cluster with active-passive failover support.