Closed nthomas-redhat closed 6 years ago
Cluster level flag set is always given priority. This is correct behaviour as per logic. If cluster level it is selected as leave-as-is then only volume level individual values are considered and if for few volumes only it's enabled cluster level it's marked as mixed.
This we had discussed in detail when we talked about enable disable profiling at volume level and then only it was coded.
@shtripat , This is not what we discussed , if I remember correctly. If that is the case what is the meaning of mixed mode? if someone wants to enable the profiling of a single volume to debug something where the global status is 'disabled', in the next sync cycle , it will be reset. I don't think this is the right behavior.
In my opinion sync should not make any auto updates during the sync. It should be purely based on the user interaction.
https://github.com/Tendrl/gluster-integration/issues/553#issuecomment-366257806, covers what I am trying to explain here.
If individual volumes to be enabled cluster import time user needs to select leave-as-is
and suring gluster integration sync based on volume profiling enabled for few volumes, cluster level volume_profiling_state
would be set as mixed
.
If while while import enable/disable
is selected every sync time cluster would try to apply the same for all the volumes. That's how it is at the moment.
After the above mentioned flow, the actions enable/disable volume profiling from cluster level should invoke the flow for setting at cluster level. Once job completes, it will take a sync to reflect all the details properly in UI.
Steps:
Result: After a while when the sync happens, the volume profiling is set as enabled
Expected result: We should not alter the profiling status during the sync cycle. Cluster level profiling status will be set to: enabled - if all volumes in the cluster are with profiling enabled disabled - if all volumes in the cluster are with profiling disbaled Mixed - if mixed configuration is found