Lack of Disk I/O Rate Goal

[imans777]

Hi,
we're using kafka cruise control in a big environment where the I/O rate of our disks are our bottleneck.
With existing goals, we can not achieve this and we have to use other goals (network and disk usages) to model a solution, but the cluster is not as optimized as expected, and we believe that a IntraBrokerDiskIORateDistributionGoal could resolve such problem.
In another word, CC should avoid putting two leader replicas on a single disk where there are multiple disks available. This could also be named IntraBrokerLeaderReplicaDistributionGoal, which I think better describes it.
Thanks.