-
-
Notifications
You must be signed in to change notification settings - Fork 13
A look at compression algorithms on the topic of Quill filesystems
chloe-everhart edited this page Jul 5, 2024
·
1 revision
Those are algorithms available on Quill ( probably )
lz4 is too weak so i didn't cared at all
compression | size |
---|---|
gzip, 128K | 69M |
lzo, 128K | 76M |
xz, 128K | 59M |
gzip, 1048576 | 69M |
lzo, 1048576 | 75M |
xz, 1048576 | 54M |
xz is the winner, so now:
56M ../out/rootfs-xz-1M-25p.squashfs
55M ../out/rootfs-xz-1M-50p.squashfs
54M ../out/rootfs-xz-1M-75p.squashfs
54M ../out/rootfs-xz-1M-100p.squashfs
nothing changed... maybe decompression speed?
averages, kindof:
100% - 1,70s
75% - 1,70s
50% - 1,70s
25% - 1,73s?
I'm wasting my time.
gzip is 0.50s
well, conclusions:
- nicolas has choosen the compression and block size wisely
- I wasted my time a bit
- use xz for max space, use gzip to be a bit faster ( if you fit it ) and no compression for the speed
In xz there is also:
-Xbcj filter1,filter2,...,filterN
Compress using filter1,filter2,...,filterN in turn (in addition to no filter), and choose the best compression. Available filters: x86, arm, armthumb, powerpc, sparc, ia64.
It doesn't do anything in Quill, at least for small archive that I have tested it against