Messages in this thread | | | Date | Sun, 9 May 2004 13:11:22 -0500 | From | Matt Mackall <> | Subject | Re: dentry bloat. |
| |
On Sun, May 09, 2004 at 08:35:55AM -0700, Linus Torvalds wrote: > > > On Sun, 9 May 2004, Guennadi Liakhovetski wrote: > > > On Sat, 8 May 2004, Linus Torvalds wrote: > > > > > 1: 5.04 % ( 5.04 % cum -- 2246) > > > 2: 5.19 % ( 10.23 % cum -- 2312) > > > > Ok, risking to state the obvious - it was intentional to count "."s and > > ".."s, wasn't it? Just this makes it a bit non-trivial to compare this > > statistics with Andrew's. > > Ok, here's a version that doesn't count "." and "..". My numbers don't > really change much for the kernel: > > and we've reached over 90% coverage with the 24-byte inline name.
I hacked up something and ran it on my webserver (which has something like 200 shell accounts). My histogram peaked at 12 rather than 10 but still hit 90% cumulative at 21 characters.
I suspect worst case is a large LAN fileserver with Samba shares and whatnot, I suspect there are a large number of "A picture of my cute puppy I took last summer.JPG" style filenames there. Anyone have stats for such an FS?
-- Matt Mackall : http://www.selenic.com : Linux development and consulting - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |