Where can a workspace administrator view the timeline of cluster resizing events to evaluate whether scaling is triggered by concurrent users or resource-intensive queries in a shared autoscaling interactive cluster?