| 1 |
cardoe |
1.1 |
GLEP: 61
|
| 2 |
|
|
Title: Manifest2 compression
|
| 3 |
robbat2 |
1.5 |
Version: $Revision: 1.4 $
|
| 4 |
|
|
Last-Modified: $Date: 2010/01/13 03:26:53 $
|
| 5 |
cardoe |
1.1 |
Author: Robin Hugh Johnson <robbat2@gentoo.org>
|
| 6 |
|
|
Status: Draft
|
| 7 |
|
|
Type: Standards Track
|
| 8 |
|
|
Content-Type: text/x-rst
|
| 9 |
|
|
Requires: 44
|
| 10 |
|
|
Created: July 2008
|
| 11 |
robbat2 |
1.3 |
Updated: October 2008, January 2010
|
| 12 |
cardoe |
1.1 |
Updates: 44
|
| 13 |
robbat2 |
1.5 |
Post-History: December 2009, January 2010
|
| 14 |
cardoe |
1.1 |
|
| 15 |
|
|
Abstract
|
| 16 |
|
|
========
|
| 17 |
|
|
Deals with compression of large Manifest2 files.
|
| 18 |
|
|
|
| 19 |
|
|
Motivation
|
| 20 |
|
|
==========
|
| 21 |
|
|
With the introduction of MetaManifest, and full-tree Manifest coverage,
|
| 22 |
|
|
we are faced with the possibility of having very large Manifests.
|
| 23 |
|
|
|
| 24 |
|
|
Preliminary experiments with MetaManifest, show that with just the
|
| 25 |
robbat2 |
1.5 |
existing per-package Manifests, the full MetaManifest (top-level only,
|
| 26 |
|
|
no first-level sub directories), for a tree including metadata/, exceeds
|
| 27 |
|
|
8MiB in size. Applying common compression can achieve a 50-60% reduction
|
| 28 |
|
|
in this size.
|
| 29 |
|
|
|
| 30 |
|
|
Additionally, some of the larger already-existing Manifests in the tree
|
| 31 |
|
|
can also be reduced.
|
| 32 |
cardoe |
1.1 |
|
| 33 |
|
|
Specification
|
| 34 |
|
|
=============
|
| 35 |
robbat2 |
1.5 |
Creation of compressed Manifests:
|
| 36 |
|
|
---------------------------------
|
| 37 |
|
|
32KiB is suggested as a arbitrary cut-off point to start generating
|
| 38 |
|
|
compressed Manifest2 files.
|
| 39 |
|
|
|
| 40 |
|
|
The compression must only applied during the creation of a tree intended
|
| 41 |
|
|
for end users. No Manifests stored in a VCS should be compressed in the
|
| 42 |
|
|
VCS. For the main gentoo-portage tree, this means that the compressed
|
| 43 |
|
|
Manifests should be generated using the CVS to Rsync process.
|
| 44 |
|
|
|
| 45 |
|
|
The Manifest compression process is required to ensure that inconsistent
|
| 46 |
|
|
compressed versions do not exist.
|
| 47 |
|
|
|
| 48 |
|
|
Validation of Manifests:
|
| 49 |
|
|
------------------------
|
| 50 |
cardoe |
1.1 |
When searching for a Manifest2 file, if the basename form does not
|
| 51 |
|
|
exist, the package manager should search in the same location using
|
| 52 |
|
|
common compressed suffixes, and use the compressed file in place of the
|
| 53 |
|
|
Manifest2.
|
| 54 |
|
|
|
| 55 |
robbat2 |
1.3 |
gzip, bzip2, lzma, xz should all be supported if available on the given
|
| 56 |
cardoe |
1.1 |
platform. In the case that multiple versions exist, the package manager
|
| 57 |
|
|
should simply pick one - they should be identical, differing only in
|
| 58 |
|
|
compression.
|
| 59 |
|
|
|
| 60 |
robbat2 |
1.5 |
Example Results with a 32KiB cut-off, gzip algorithm
|
| 61 |
|
|
====================================================
|
| 62 |
|
|
As of 2010/01/30, the suggested cut-off would impact the following 21 existing
|
| 63 |
|
|
Manifests, for a saving of nearly 900KiB::
|
| 64 |
|
|
|
| 65 |
|
|
Size Path
|
| 66 |
|
|
65788 app-doc/linux-gazette/Manifest
|
| 67 |
|
|
75739 app-office/openoffice-bin/Manifest
|
| 68 |
|
|
40534 app-text/texlive-core/Manifest
|
| 69 |
|
|
41710 dev-texlive/texlive-bibtexextra/Manifest
|
| 70 |
|
|
38197 dev-texlive/texlive-documentation-english/Manifest
|
| 71 |
|
|
129610 dev-texlive/texlive-fontsextra/Manifest
|
| 72 |
|
|
36022 dev-texlive/texlive-humanities/Manifest
|
| 73 |
|
|
686118 dev-texlive/texlive-latexextra/Manifest
|
| 74 |
|
|
43392 dev-texlive/texlive-latexrecommended/Manifest
|
| 75 |
|
|
33375 dev-texlive/texlive-mathextra/Manifest
|
| 76 |
|
|
39781 dev-texlive/texlive-pictures/Manifest
|
| 77 |
|
|
69567 dev-texlive/texlive-pstricks/Manifest
|
| 78 |
|
|
75460 dev-texlive/texlive-publishers/Manifest
|
| 79 |
|
|
50879 dev-texlive/texlive-science/Manifest
|
| 80 |
|
|
36711 kde-base/kde-l10n/Manifest
|
| 81 |
|
|
36539 media-gfx/bootsplash-themes/Manifest
|
| 82 |
|
|
33058 net-fs/autofs/Manifest
|
| 83 |
|
|
39781 www-client/firefox-bin/Manifest
|
| 84 |
|
|
48983 www-client/icecat/Manifest
|
| 85 |
|
|
60213 www-client/mozilla-firefox/Manifest
|
| 86 |
|
|
39065 x11-themes/gkrellm-themes/Manifest
|
| 87 |
|
|
|
| 88 |
|
|
|
| 89 |
|
|
Additionally, with the MetaManifest proposal, the following new manifests would
|
| 90 |
|
|
also be compressed, for a saving of nearly 4MiB::
|
| 91 |
|
|
|
| 92 |
|
|
Size Path
|
| 93 |
|
|
33442 app-admin/Manifest
|
| 94 |
|
|
71073 app-dicts/Manifest
|
| 95 |
|
|
35923 app-emacs/Manifest
|
| 96 |
|
|
45808 app-misc/Manifest
|
| 97 |
|
|
50169 app-text/Manifest
|
| 98 |
|
|
112786 dev-java/Manifest
|
| 99 |
|
|
65581 dev-libs/Manifest
|
| 100 |
|
|
42619 dev-lisp/Manifest
|
| 101 |
|
|
182163 dev-perl/Manifest
|
| 102 |
|
|
96198 dev-python/Manifest
|
| 103 |
|
|
58963 dev-ruby/Manifest
|
| 104 |
|
|
59736 dev-util/Manifest
|
| 105 |
|
|
58338 eclass/Manifest
|
| 106 |
|
|
55749 kde-base/Manifest
|
| 107 |
|
|
110064 licenses/Manifest
|
| 108 |
|
|
35262 media-gfx/Manifest
|
| 109 |
|
|
53995 media-libs/Manifest
|
| 110 |
|
|
55607 media-plugins/Manifest
|
| 111 |
|
|
71911 media-sound/Manifest
|
| 112 |
|
|
34835 media-video/Manifest
|
| 113 |
|
|
5747849 metadata/Manifest
|
| 114 |
|
|
47452 net-analyzer/Manifest
|
| 115 |
|
|
65989 net-misc/Manifest
|
| 116 |
|
|
316787 profiles/Manifest
|
| 117 |
|
|
67784 sys-apps/Manifest
|
| 118 |
|
|
48971 x11-misc/Manifest
|
| 119 |
|
|
41475 x11-plugins/Manifest
|
| 120 |
|
|
|
| 121 |
cardoe |
1.1 |
|
| 122 |
|
|
Backwards Compatibility
|
| 123 |
|
|
=======================
|
| 124 |
|
|
The package Manifests should also be maintained as ONLY uncompressed in
|
| 125 |
|
|
CVS.
|
| 126 |
|
|
|
| 127 |
|
|
For processing of all existing per-package Manifests, if compression is
|
| 128 |
|
|
used, it should be done in parallel to the existing Manifests, to
|
| 129 |
|
|
provide for a changeover period. Newer versions of Portage may later
|
| 130 |
|
|
choose to exclude all non-compressed Manifests during emerge --sync if
|
| 131 |
|
|
compressed versions are guaranteed to exist on the servers.
|
| 132 |
|
|
|
| 133 |
|
|
MetaManifests may come into existence as compressed from the start, as
|
| 134 |
|
|
do not have an backwards compatibility issues.
|
| 135 |
|
|
|
| 136 |
|
|
As a side note, this breaks all manual interaction with Manifests
|
| 137 |
|
|
such as grep, and so should only be applied to large Manifest2 files,
|
| 138 |
robbat2 |
1.5 |
such as the MetaManifest.
|
| 139 |
cardoe |
1.1 |
|
| 140 |
|
|
References
|
| 141 |
|
|
==========
|
| 142 |
|
|
.. [#GLEP44] Mauch, M. (2005) GLEP44 - Manifest2 format.
|
| 143 |
|
|
http://www.gentoo.org/proj/en/glep/glep-0044.html
|
| 144 |
|
|
|
| 145 |
|
|
Copyright
|
| 146 |
|
|
=========
|
| 147 |
robbat2 |
1.4 |
Copyright (c) 2008-2010 by Robin Hugh Johnson. This material may be
|
| 148 |
cardoe |
1.1 |
distributed only subject to the terms and conditions set forth in the
|
| 149 |
|
|
Open Publication License, v1.0.
|
| 150 |
|
|
|
| 151 |
|
|
vim: tw=72 ts=2 expandtab:
|