This counter counts page faults resolved cheaply by reclaiming a page still sitting on the free-page list rather than reading from disk. A high proportion of faults resolved this way, compared to faults requiring an actual disk read, is a good sign, since it means the system is absorbing memory pressure without expensive disk I/O. For example, if most of an interval's page faults show up here rather than as disk reads, the admin can conclude the free list is currently large enough to cushion memory demand.