On Mon, 28 Jun 1999 12:13:15 -0500 (EST), "Mark H. Wood"
<mwood@IUPUI.Edu> said:
> People who run VMS systems become all too familiar with the
> characteristics of extent mapping. Files-11 tends to cause
> slowly-expanding files to wind up with lots of short extents if the
> filesystem is nearing capacity. There's a healthy market for
> defragmentation products in this space.
There are two _entirely_ different matters here: the algorithm by
which you allocate new blocks, and the way in which you encode the
allocation mapping information on disk. You can use extent mapping
without touching the allocation algorithm in any way whatsoever.
Extent mapping, by itself, will not affect the fragmentation of the
filesystem in any significant way whatsoever (the only impact it can
possibly have is over the allocation of the mapping blocks themselves,
and by reducing the volume of mapping information you actually expect
extent maps to improve things here).
> No doubt this was one of the things that prompted the development of
> Spiralog.
Actually, LFS filesystems introduce an entirely new class of fragmentation
problems. :)
--Stephen
-
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/