Gentoo Archives: gentoo-dev

From: Todd Berman <tberman@g.o>
To: Andrew Gaffney <agaffney@×××××××××××.com>
Cc: gentoo-dev@l.g.o
Subject: Re: [gentoo-dev] Idea for the portage maintainers
Date: Mon, 12 Apr 2004 17:42:34
Message-Id: 1081791587.2075.15.camel@localhost.localdomain
In Reply to: Re: [gentoo-dev] Idea for the portage maintainers by Andrew Gaffney
1 On Mon, 2004-12-04 at 12:17 -0500, Andrew Gaffney wrote:
2 > Todd Berman wrote:
3 > > On Mon, 2004-12-04 at 11:59 -0500, Andrew Gaffney wrote:
4 > >
5 > >>Todd Berman wrote:
6 > >>
7 > >>>On Mon, 2004-12-04 at 11:22 -0500, Andrew Gaffney wrote:
8 > >>>
9 > >>>
10 > >>>>Troy Dack wrote:
11 > >>>>
12 > >>>>
13 > >>>>>Another point against a monolithic zip containing all the ebuilds (or
14 > >>>>>even per directory zips) is the performance hit that slow machines would
15 > >>>>>take, not everybody runs gentoo on a 2GHz plus machine (eg: my little
16 > >>>>>PII-400 in the corner)
17 > >>>>
18 > >>>>Or my little P233 Thinkpad...
19 > >>>>
20 > >>>
21 > >>>
22 > >>>And with the current setup of writing thousands of 1K files that little
23 > >>>p233 thinkpad really flys i bet...
24 > >>
25 > >>I'm not sure if that was supposed to be sarcastic, but yes, it does fly. It only takes
26 > >>slightly longer to sync than my Athlon 1.3GHz desktop. The only part that takes forever is
27 > >>updating the portage cache. That's why I just use a NFS shared portage tree from my
28 > >>desktop machine now.
29 > >
30 > > In a way it was and in a way it wasn't. I honestly don't understand how
31 > > you can explain to me that a compression-less zip file be any slower
32 > > than the current setup.
33 > >
34 > > With compression I could understand, but without I don't think any speed
35 > > difference would be noticeable. However, even with compression the
36 > > operation should be fairly fast. zip is not like a tar.gz or tar.bz2,
37 > > ie, you can read a file out of it and search through it fairly fast, and
38 > > you don't have to uncompress the entire archive to get a single file.
39 >
40 > Even without compression, there is still a little bit of overhead with having ebuilds and
41 > such contained in ZIP files. The overhead comes into play both during sync and emerge. It
42 > wouldn't be noticable on a fast machine (e.g. my Athlon 1.3GHz desktop) but it would make
43 > an operation like 'emerge -uDpv world' take quite a bit longer on my Thinkpad.
44 >
45
46 This is going to be a slow operation on that thinkpad regardless.
47
48 > As someone else pointed out, the ZIP files wouldn't make syncs faster either. Rsync
49 > transfers only the differences in files and not the entire app-text directory or whatever
50 > category you're working with.
51 >
52
53 Portage does not use this as it takes more cpu server and client side to
54 figure out what changed with 80 thousand 1K files than it does to just
55 transfer the files across.
56
57 > As another person pointed out, it would be more difficult to modify ebuilds in the tree by
58 > hand when they're contained in ZIP files.
59 >
60
61 This could easily be dealt with. the cvs->rsync process could create zip
62 files, and portage would use the zip stuff by default for /usr/portage/
63 and use the old directory structure code for your overlays.
64
65 > It was a good idea, but I don't think it is practical.
66
67 It is absolutely practical, and I think would cut down on the time taken
68 to sync (because now you could turn compression back on, and actually
69 transfer just changes because you wouldnt be syncing 80 thousand files).
70 It might potentially have small speed loses when running emerge -vuDp
71 world, but as that process takes a long time regardless, I don't think
72 it would be noticeable. The time it takes to read a single file out of
73 an uncompressed zip is not noticeably longer than the time it takes to
74 read a single file off the hard drive.
75
76 Note, as of writing this the portage tree takes up 333MB of space on my
77 hard drive and a compression-less zipfile of the tree takes up 74MB. To
78 me, the potential savings are incredible.
79
80 --Todd
81
82
83 --
84 gentoo-dev@g.o mailing list

Replies

Subject Author
Re: [gentoo-dev] Idea for the portage maintainers Jason Stubbs <jstubbs@g.o>