| 1 |
On Thu, 10 May 2007 11:34:46 +0100 |
| 2 |
Neil Bothwick <neil@××××××××××.uk> wrote: |
| 3 |
|
| 4 |
> On Thu, 10 May 2007 12:11:34 +0200, Benno Schulenberg wrote: |
| 5 |
> |
| 6 |
> > > No idea, but I tried it when I encountered that page and portage |
| 7 |
> > > operations were measurably faster. |
| 8 |
> > |
| 9 |
> > That might well be just the transfer effect: you went from an old |
| 10 |
> > fragmented file system to a fresh unfragmented one. |
| 11 |
> |
| 12 |
> I allowed for that. I created a new filesystem for /usr/portage - I |
| 13 |
> had been using a directory in /usr before. |
| 14 |
> |
| 15 |
> |
| 16 |
Well, maybe it has to do with the efficiency of reading discontiguous |
| 17 |
blocks from one file as opposed to a filesystem. Since it's a sparse |
| 18 |
file, there might be a lot of 'space' that, if it were on an actual |
| 19 |
disk, the heads would have to pass over; thus there may be a way in |
| 20 |
which a sparse file is more efficient than a regular filesystem. |
| 21 |
|
| 22 |
Remeber that the files in portage are, except for distfiles, quite |
| 23 |
small. By my calculation, the average size for files and directories |
| 24 |
under $PORTDIR (excluding $DISTDIR of course) is only 62 bytes. What |
| 25 |
would you bet that on a disk partition, the other 962 to 4034 bytes per |
| 26 |
block (I couldn't have block sizes less than 1K on reiser for my |
| 27 |
portage, and 4096 is the default for most FS's) are filled with |
| 28 |
nothing, and the heads need to pass over them to read the next block. |
| 29 |
On a sparse file that space is merely reserved, it needn't actually |
| 30 |
exist. Hope that helps you conceptualize the difference. |
| 31 |
-- |
| 32 |
gentoo-user@g.o mailing list |
Archives