I have this issue as well. I just over-allocate nodes to my cluster so every node is comfortably below a disk usage threshold rather than waste time on optimizi

I have this issue as well. I just over-allocate nodes to my cluster so every node is comfortably below a disk usage threshold rather than waste time on optimizing this.