Messages in this thread | | | From | Heinz Mauelshagen <> | Subject | Re: [patch] dynamic buffer cache hash table size | Date | Sat, 23 Jan 1999 3:14:30 MET |
| |
> > On Fri, 22 Jan 1999, Heinz Mauelshagen wrote: > > > The bad performance nearly stays the same, when the buffermem is large 8*( > > Only adding my artificial limitation for the buffermem to limit it > > to for eg. 20M is almost fine. > > Ok, now please do a crazy real hack test. > > I want you alloc 1000 pages (4mbyte of RAM) for the buffer cache hash. If > also that way you won't see differences I will be interested in other > test. > > Please start with the last patch applyed. You'll find a function called > my new buffer_hash_init() near line 1500 of linux/fs/buffers.c > > Plese edit this code: > > mem_pages = (mem_pages + (1<<9)) >> 9; > > and modify it to: > > mem_pages = 1000; > > Then recompile and reboot. If you'll still see bad performances with the > memory full of buffers could you please append="profile=4" in lilo.conf? > > Then reboot and tell me which are the most run functions (maybe > shrink_mmap?). >
Did hack and recompile/reboot/test. Performance stays bad. readprofile shows the same results i've sent already.
Regards, Heinz
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Systemmanagement Entwicklungsbereich 2 Deutsche Telekom AG Entwicklungszentrum Darmstadt Heinz Mauelshagen Otto-Roehm-Strasse 71c Postfach 10 05 41 mge@ez-darmstadt.telekom.de 64205 Darmstadt Germany +49 6151 886-425 FAX-386 =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Systemmanagement Entwicklungsbereich 2 Deutsche Telekom AG Entwicklungszentrum Darmstadt Heinz Mauelshagen Otto-Roehm-Strasse 71c Postfach 10 05 41 mge@ez-darmstadt.telekom.de 64205 Darmstadt Germany +49 6151 886-425 FAX-386 =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |