Proposal for solving CD Size problems

[Vassilis Pandis]

--- Scott James Remnant <scott at ubuntu.com> wrote:


> I've done some tests with this; the results aren't great:
> 
> -rw-r--r-- 1 root root 599M 2006-09-27 20:23 filesystem.squashfs_gz
> -rw-r--r-- 1 root root 552M 2006-09-27 20:37 filesystem.squashfs_7z
> 
> 
> The saving is under 8%, and the performance hit is extremely noticeable
> (at least within VMware), it takes almost twice the time to boot.
> 

I grabbed yesterday's daily live and tried similar benchmarks (both with squashfs3.0):

I used patches from : ftp://ftp.slax.org/SLAX-6.x/testing/technology-preview/devel/

 - Edgy's filesystem.squashfs is 623MB. LZMA gives 548MB . 

 - I tried a couple of terribly unscientific benchmarks: I ran unsquashfs a couple of times
to make sure data is loaded into cache and then ran it 3 times on the lzma and zlib squashfses

Results: 1m55s for LZMA , 22s for zlib. I removed most of the code in unsquashfs that writes to
the disk to make sure that disk I/O is less of an issue. 

My results appear more or less in agreement with Scott's (12 % savings instead of 8%). 



		
___________________________________________________________ 
Copy addresses and emails from any email account to Yahoo! Mail - quick, easy and free. http://uk.docs.yahoo.com/trueswitch2.html