1 |
On Tue, 13 May 2014 08:18:25 -0400 Rich Freeman wrote: |
2 |
> On Tue, May 13, 2014 at 7:01 AM, Andrew Savchenko <bircoph@×××××.com> wrote: |
3 |
> > |
4 |
> > If we are trying to consider all possible cases, some filesystems |
5 |
> > may benefit even from compression of very small files (e.g. from |
6 |
> > 140 to 100 bytes) due to packing of multiple small files in the |
7 |
> > same inode. ReiserFS is a good example, but more may be somewhere |
8 |
> > there. |
9 |
> > |
10 |
> |
11 |
> Btrfs also supports file inlining, so every byte saved on small files |
12 |
> does actually help (I believe the data structure that stores the |
13 |
> inlined data doesn't have a fixed record size). Then again, btrfs |
14 |
> also supports lzo compression and I believe this is fairly widely |
15 |
> used, so I'm not sure that the impact of not compressing small files |
16 |
> will be felt. |
17 |
|
18 |
I did not meant inlining. I was talking about block suballocation |
19 |
which allows to store small files in underused blocks of another |
20 |
files: |
21 |
http://en.wikipedia.org/wiki/Block_suballocation |
22 |
|
23 |
> I don't think ext4 supports inlining, but I see some discussions of |
24 |
> attempts to add it. |
25 |
|
26 |
Ext4 supports inlining for files up to 59 bytes: |
27 |
https://ext4.wiki.kernel.org/index.php/Ext4_Disk_Layout#Inline_Data |
28 |
|
29 |
> For VERY small files I would think that overhead would become an issue. |
30 |
> |
31 |
> Unless we have a bunch of 30-byte man pages I'd think that both |
32 |
> simplicity and some potential for utility would lead us to use the |
33 |
> best algorithm possible. |
34 |
|
35 |
Agreed, though performance should be considered still. I doubt |
36 |
paq8l -9 will be used for this task, though it is about 1.5 times |
37 |
more effective than xz -9e on text files, even on small ones like |
38 |
man pages; on large files it is at least 2 times better. |
39 |
|
40 |
Best regards, |
41 |
Andrew Savchenko |