Date: Fri, 26 Sep 1997 02:36:30 +1000 (EST) From: Darren Reed <darrenr@cyber.com.au> To: phk@FreeBSD.ORG (Poul-Henning Kamp) Cc: cvs-committers@FreeBSD.ORG, cvs-all@FreeBSD.ORG, cvs-sys@FreeBSD.ORG Subject: Re: cvs commit: src/sys/kern vfs_subr.c Message-ID: <199709251636.CAA00930@plum.cyber.com.au> In-Reply-To: <199709251618.JAA22355@freefall.freebsd.org> from "Poul-Henning Kamp" at Sep 25, 97 09:18:00 am
next in thread | previous in thread | raw e-mail | index | archive | help
In some mail I received from Poul-Henning Kamp, sie wrote > > phk 1997/09/25 09:18:00 PDT > > Modified files: > sys/kern vfs_subr.c > Log: > Reduce the target number of vnodes on the freelist from desiredvnodes > (usually a couple of thousand) to 25. The measured impact on cache-hits > doesn't justify spending memory this way: > > Target number of free vnodes versus namecache hit rate in % during a > make world: > 10 98.5316 > 200 98.5479 > 500 98.5546 > 1000 98.5709 > 3000 98.6006 > 4000 98.6126 Is this always a good measure of a file system ? Sure, there maybe other reasons to do it, but to me, "make world" doesn't exactly exhibit random file system behaviour or anything but intensive use of certain parts, in a certain order for a set amount of time. All in all a very limited behaviour pattern which will cache well, as opposed to (say) 200 users connecting to a news server. I'd be more interested in some other results, if they're available...? Darren
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199709251636.CAA00930>