On Wed, Aug 07, 2024 at 02:46:29PM +0800, Yong Huang wrote: > I still think dropping the dirty_rate_high_cnt is a refinement. If migration > has a precise detection of convergence, what do you think of it?
Maybe; I don't think I thought through those yet.. If so that'll make sense to be based on the sync change first, then we can have a parameter (doesn't sound like a cap on its own) for keeping the old auto-converge, and pile new logics there. Thanks, -- Peter Xu