>> http://www.westnet.com/~gsmith/content/linux-pdflush.htm
only tangentially. I do normally tune the dirty parameters to write/clean dirty pages more agressively, but this really has nothing to do with the drop_cache behavior being advocated. >> If anyone knows of a more recent version of the RHEL paper referenced >> therein: >> >> http://people.redhat.com/nhorman/papers/rhel4_vm.pdf I read it quickly just now and think it's still reasonably accurate. obviously, 64b systems are the norm these days, which disappears the zone Highmem. the topic being discussed is really all about two edges in figure 3: using the drop_caches sledgehammer just pushes all pages from inactive-clean to free, making it impossible for inactive-clean pages to be reactivated (ie, avoiding a read.) regards, mark hahn. _______________________________________________ Beowulf mailing list, [email protected] sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf
