Issue with partition var reduce_retention_period_immediately

[eambrosyupgrade]

Problem

The variable reduce_retention_period_immediately is optional however it doesn't seem to track the changes in state.

If I reduce the retention period of a partition in terraform from say 100 > 90; and have this option set to false, it applies "successfully" but since the change is staged and the retention period is still 100d for the next week, I get continual drift in my terraform pipeline since the terraform state expects that the retention period should be 90d now, and is unaware that retention will change eventually.

So in a pipeline I have to set this to true, otherwise I get constant drift and applies that attempt to set the partition to a new retention period, in 7 days from the last apply.

Expected Behavior

The provider should record that this change is "staged"; timestamps of the swap and compare that with the timestamps for the retention reduction. Looking at the API here I think the issue is that that response value of retentionEffectiveAt isn't in state and doesn't seem to exist in the partition's resource.

[dlinsumo]

Internal jira SUMO-251300