Large filesystems with massive inodes on BSD5.2

Doug White dwhite at
Sat Jan 10 12:19:19 PST 2004

On Fri, 9 Jan 2004, Tom Arnold wrote:

> On Fri, Jan 09, 2004 at 04:56:39PM -0800, Doug White wrote:
> > Unless you're going to fill the filesystem with 512 byte files, you
> > probably have the sizes set too low.  You're probably OK with the default
> > of 4096 bytes/inode.
> Its actually gonna be filled with gzipped spam which we find seems to be
> under 1k. 512bytes is probably too small, but even doubling the fragment
> and inode size to 1024 still leaves me with a billion inodes...

This is a really inefficient filing system for this kind of data .. you
are really going to be sorry if you ever have to move it.  Can you stick
it in a database, or combine it into larger files? The gzip overhead is
going to consume a significant fraction of the space.

Doug White                    |  FreeBSD: The Power to Serve
dwhite at          |

More information about the freebsd-fs mailing list