yum のリポジトリメタデータに含まれる RPM パッケージファイルのサイズ情報などからリポジトリ容量を概算

yum-repo-size-info-list.sh

#! /bin/bash
#
# Copyright(C) 2016 Red Hat, Inc.
# Author: Satoru SATOH <[email protected]>
# License: MIT
#
# Example:
# [root@rhel-7-client-1 ~]# ./yum-repo-size-info-list.sh \
# > -r rhel-7-server-rpms -r rhel-7-server-optional-rpms \
# > -r rhel-7-server-rh-common-rpms -r rhel-7-server-extras-rpms \
# > -r rhel-7-server-supplementary-rpms -r rhel-server-rhscl-7-rpms
# Loaded plugins: changelog, product-id, subscription-manager
# repo id                                     repo name                                            status
# rhel-7-server-extras-rpms/x86_64            Red Hat Enterprise Linux 7 Server - Extras (RPMs)       188
# rhel-7-server-optional-rpms/7Server/x86_64  Red Hat Enterprise Linux 7 Server - Optional (RPMs)   8,516
# rhel-7-server-rh-common-rpms/7Server/x86_64 Red Hat Enterprise Linux 7 Server - RH Common (RPMs)    186
# rhel-7-server-rpms/7Server/x86_64           Red Hat Enterprise Linux 7 Server (RPMs)             10,697
# rhel-7-server-supplementary-rpms/7Server/x86_64
#                                             Red Hat Enterprise Linux 7 Server - Supplementary (R    129
# rhel-server-rhscl-7-rpms/7Server/x86_64     Red Hat Software Collections RPMs for Red Hat Enterp  4,594
# repolist: 24,310
# Loaded plugins: changelog, product-id, subscription-manager
# rhel-7-server-extras-rpms                                                       | 2.9 kB  00:00:00
# rhel-7-server-optional-rpms                                                     | 3.5 kB  00:00:00
# rhel-7-server-rh-common-rpms                                                    | 3.8 kB  00:00:00
# rhel-7-server-rpms                                                              | 3.7 kB  00:00:00
# rhel-7-server-supplementary-rpms                                                | 3.7 kB  00:00:00
# rhel-server-rhscl-7-rpms                                                        | 3.1 kB  00:00:00
# Metadata Cache Created
# [Info] rhel-7-server-rpms: 10697 rpms, 13658590872 bytes (rpms), 238828 bytes (metadata)
# rhel-7-server-rpms,10697,13658590872,238828,13658829700
# [Info] rhel-7-server-extras-rpms: 188 rpms, 417264100 bytes (rpms), 1532 bytes (metadata)
# rhel-7-server-extras-rpms,188,417264100,1532,417265632
# [Info] rhel-7-server-optional-rpms: 8516 rpms, 11226677420 bytes (rpms), 103260 bytes (metadata)
# rhel-7-server-optional-rpms,8516,11226677420,103260,11226780680
# [Info] rhel-7-server-rh-common-rpms: 186 rpms, 4807985928 bytes (rpms), 1508 bytes (metadata)
# rhel-7-server-rh-common-rpms,186,4807985928,1508,4807987436
# [Info] rhel-7-server-supplementary-rpms: 129 rpms, 3418212204 bytes (rpms), 3872 bytes (metadata)
# rhel-7-server-supplementary-rpms,129,3418212204,3872,3418216076
# [Info] rhel-server-rhscl-7-rpms: 4594 rpms, 4004597412 bytes (rpms), 68988 bytes (metadata)
# rhel-server-rhscl-7-rpms,4594,4004597412,68988,4004666400
# All repo data size: 37533745924 bytes = 35794 MB = 34 GB
# [root@rhel-7-client-1 ~]#

set -e

REPOS=""
YUM_REPO_DIR="/var/cache/yum"
YUM_REPO_ENABLE_OPTS=""

usage () {
    cat << EOU
Usage: $0 [Options]

Options:
  -r REPO   Specify yum repo ID[s] to limit yum repos to get result.
            It can be given multiple times. 

Example:
  $0 -r rhel-7-server-rpms -r rhel-7-server-optional-rpms
EOU
    exit 0
}

while getopts "r:hv" OPT; do
    case $OPT in
      r) REPOS="$REPOS $OPTARG"
         YUM_REPO_ENABLE_OPTS="${YUM_REPO_ENABLE_OPTS} --enablerepo=$OPTARG";;
      \?|h) usage ;;
    esac
done
shift $(( $OPTIND - 1 ))

# Primary DB path Examples:
# rhel-5: /var/cache/yum/*/*-primary*.sqlite
# rhel-6: /var/cache/yum/x86_64/6Server/*/*-primary*.sqlite
# rhel-7: /var/cache/yum/x86_64/7Server/*/gen/primary_db.sqlite
sum_of_data_size_min=0

if test "x${YUM_REPO_ENABLE_OPTS}" = "x"; then
    yum repolist
    yum makecache
else
    yum repolist --disablerepo='*' ${YUM_REPO_ENABLE_OPTS}
    yum makecache --disablerepo='*' ${YUM_REPO_ENABLE_OPTS}
fi

for repo_pkgdb in $(find ${YUM_REPO_DIR} -regex '.*primary.*\.sqlite$')
do
    if test -d /etc/systemd; then
        repo_name=${repo_pkgdb/\/gen*/}
        repo_name=${repo_name##*/}
    else
        repo_name=$(echo ${repo_pkgdb} | sed -r "s,${YUM_REPO_DIR}/.+/.+/([^/]+)/.*,\1,")
    fi

    if test "x${REPOS}" != "x"; then
        found=0
        for repo in $REPOS; do
            test "x${repo}" = "x${repo_name}" && found=1 || :
        done
        test ${found} -eq 1 || continue
    fi
    repo_num_of_rpms=$(sqlite3 ${repo_pkgdb} "select count(name) from packages")
    repo_sum_of_rpmsize=$(sqlite3 ${repo_pkgdb} "select sum(size_package) from packages")
    repo_metadata_size=$(du -s ${repo_pkgdb%/*} | sed -r 's/[[:blank:]].*//g')
    repo_data_size_min=$(( ${repo_sum_of_rpmsize} + ${repo_metadata_size} ))
    sum_of_data_size_min=$(( ${sum_of_data_size_min} + ${repo_data_size_min} ))

    echo "[Info] ${repo_name}: ${repo_num_of_rpms} rpms, ${repo_sum_of_rpmsize} bytes (rpms), ${repo_metadata_size} bytes (metadata)"
    echo "${repo_name},${repo_num_of_rpms},${repo_sum_of_rpmsize},${repo_metadata_size},${repo_data_size_min}"
done

echo "All repo data size: ${sum_of_data_size_min} bytes = $(( ${sum_of_data_size_min} / 1024 / 1024 )) MB = $(( ${sum_of_data_size_min} / 1024 / 1024 / 1024 )) GB"

# vim:sw=4:ts=4:et:

Maybe I should extend it to get size data from primary.xml.gz:

<?xml version="1.0" encoding="UTF-8"?>
<metadata packages="10675" xmlns="http://linux.duke.edu/metadata/common" xmlns:rpm="http://linux.duke.edu/metadata/rpm"><package type="rpm">
  <name>at-sysvinit</name>
             ...
  <size package="6440" installed="2122" archive="2380"/><!-- size[@package] is size of the RPM package file -->
...