Can swap-in be improved?

Posted May 25, 2007 1:34 UTC (Fri) by interalia (subscriber, #26615)
In reply to: Can swap-in be improved? by NAR
Parent article: On-demand readahead

A plausible culprit is the cronjob that runs updatedb so that your 'locate' database is up-to-date. That churns the cache quite a bit since it involves a lot of disk seeking.