deadline seems less useful now that a bunch of edge cases were fixed in cfq (the graphs used to look way worse than this): http://blog.pgaddict.com/posts/postgr

deadline seems less useful now that a bunch of edge cases were fixed in cfq (the graphs used to look way worse than this): http://blog.pgaddict.com/posts/postgresql-io-schedulers-cfq-noop-deadline http://blog.pgaddict.com/posts/postgresql-io-schedulers-cfq-...