Shulou(Shulou.com)06/03 Report--

Part two of Linux package Management

Front-end tools for package management:

CentOS system, yum,dnf

Yum:

Yellowdog Updater Modified

Yellowdog is a distribution linux that uses rpm as the default package management tool

URL: the main identification method of yum positioning software repository and package files in the repository.

Schema:// [username@] hostname [: port] [/ PATH/TO/DOCUMENT] [parameters] [method]

Web site:

Http/https

Http://www.163.com

Yum is the model of Candace S architecture, in which the server is also called software repository (yum repository), and the client needs to write URL to specify the server and package path.

On the server, the software repository is actually a directory containing package files and metadata files

Metadata files are uniformly stored in the repodata directory; in other words, the path where the repodata directory is located is the path to the software repository

Createrepo command:

Createrepo-Create repomd (xml-rpm-metadata) repository

Createrepo [options]

The directory where the package is stored

Use of the client:

Yum package

Main configuration file for yum: / etc/yum.conf

Ini-style configuration information mode: segmented configuration content

[repositoryid]

Statements

Note: all configuration statements take effect from one [] to the next []

For example:

[repo1]

Statements

[repo2]

Statements

Note:

1.repositoryid must be globally unique

2.repositoryid cannot be included except. Any special character other than _, can contain letters and numbers, and letters are strictly case-sensitive

Contents of / etc/yum.conf file

[main]

Cachedir=/var/cache/yum/$basearch/$releasever

Keepcache=0

Debuglevel=2

Logfile=/var/log/yum.log

Exactarch=1

Obsoletes=1

Gpgcheck=1

Plugins=1

Installonly_limit=5

Bugtracker_url= http://bugs.centos.org/set_project.php?project_id=23&ref=http://bugs.centos.org/bug_report_page.php?category=yum

Distroverpkg=centos-release

For more comprehensive information, you can get it using man 5 yum.conf

# PUT YOUR REPOS HERE OR IN separate files named file.repo in / etc/yum.repos.d

The pointing information of user-defined repository can be written in the main configuration file, or the pointing information of user-defined repository can be suffixed with .repo and stored in the / etc/yum.repos.d directory.

Master profile:

/ etc/yum.conf

Only the public configuration content that provides yum repository pointing information

Note: in general, no changes are made to this file.

Fragment profile:

/ etc/yum.repos.d/*.repo

Each file with the suffix .repo is a repository file; so each file can be configured to point to one or more yum repositories

Yum programs look for yum repositories according to these configuration files; each time they may find multiple yum repositories, find the latest version of packages from all available repositories and cache them locally for installation, upgrade and other operations

Multiple repositories of the same kind are called mirrors; as long as one is available

There can also be multiple repositories of different types, and yum can use these repositories at the same time.

How should the .repo file be defined:

[repositoryid]

Name=Full name of repository

Baseurl=schema1://hostname/PATH/TO/REPODATA

Schema2://hostname/PATH/TO/REPODATA

Schema3://hostname/PATH/TO/REPODATA

Schema4://hostname/PATH/TO/REPODATA

Schema:

Http [s], ftp, nfs: the network server points to

File: the local server points to

# mirrorlist=mirror_url

Enabled= {0 | 1}. 0: disable this warehouse pointing; 1: enable; if there is no enabled statement, it is enabled by default

Gpgcheck= {0 | 1}. 0 means no package authentication. 1 means authentication. If 1, you need to specify the public key file required for authentication and use the gpgkey statement.

Gpgkey=schema://hostname/PATH/TO/PUBLIC_KEY_FILE

Example: gpgkey= file:///etc/pki/rpm-gpg/RPM-GPG-KEY-CentOS-$releasever

Failovermethod= {roundrobin | priority}

Fault switching method:

Roundrobin: random search, default method

Priority: select in order

Enablegroups= {1 | 0}

Cost=num, the default is 1000 for all warehouses.

A general repo points to the content:

[repoid]

Name=

Baseurl=

# mirrorlist=

Enbaled=1

Gpgcheck=0

# gpgkey=

Usage of the yum command:

Yum-Yellowdog Updater Modified

Yum [options] [command] [package...]

[options]:

-y,-- assumeyes: all questions that interact with users during yum execution are answered in yes

-Q,-- quiet: quiet mode, during a certain yum execution, no information is output to stdout

-- enablerepo=repoidglob

-- disablerepo=repoidglob

During a certain yum execution, it is explicitly specified to enable or disable a repository point; after using such an option, the function of the warehouse pointing to the enabled statement in the file will be invalidated.

-- noplugins:

During a certain yum execution, make it clear that no plug-ins are used

-- nogpgcheck:

During a certain yum execution, it is clearly stated that if the integrity of the package is not verified, the gpgcheck=1 statement in the repository pointing to the file will fail

[command]

Command is one of:

* install package1 [package2] [...]

Install the package

Note:

1. If you just give the package name, yum searches all repositories and finds the latest version of the cache installation

two。 If the name and version of the package are given, look for the package that installs the specified version

3. If the package given is the full package name, you are only looking for a package

4. The package name can be globbing, (*? [] [^])

* update [package1] [package2] [...]

* update-to [package1] [package2] [...]

* check-update

* upgrade [package1] [package2] [...]

* upgrade-to [package1] [package2] [...]

* downgrade package1 [package2] [...]

Package upgrade or downgrade:

Alternation between different versions of a package

* remove | erase package1 [package2] [...]

Uninstall the package:

* list [...]

Yum list LIST OPTIONS

LIST OPTIONS:

All: displays all packages, including those that have been installed and those that are not yet installed but available

Note:

The installer of the anaconda:CentOS system, the package that is installed when installing the operating system

Installed: displays all installed packages

Available: displays all packages that have not been installed but can be installed

Updates: shows all packages that can be upgraded and updated

* info [...]

Yum info LIST OPTIONS

LIST OPTIONS:

All: displays all packages, including those that have been installed and those that are not yet installed but available

Note:

The installer of the anaconda:CentOS system, the package that is installed when installing the operating system

Installed: displays all installed packages

Available: displays all packages that have not been installed but can be installed

Updates: shows all packages that can be upgraded and updated

Name [globbing]

* provides | whatprovides feature1 [feature2] [...]

See which package provides the specified feature

* clean [packages | metadata | expire-cache | rpmdb | plugins | all]

Clean up local cache information

* makecache

Build cach

* groupinstall group1 [group2] [...]

* groupupdate group1 [group2] [...]

* grouplist [hidden] [groupwildcard] [...]

* groupremove group1 [group2] [...]

* groupinfo group1 [...]

Administrative commands for the package group:

* search string1 [string2] [...]

Match by keyword, by default, search for package name and summary information

Fuzzy search

* shell [filename]

Provide yum with an interactive command interface

* localinstall rpmfile1 [rpmfile2] [...]

(maintained for legacy reasons only-use install)

* localupdate rpmfile1 [rpmfile2] [...]

(maintained for legacy reasons only-use update)

Implement local package installation, that is, operate rpm format package files directly without relying on the software repository

These two commands have been deactivated and replaced with install and update

* reinstall package1 [package2] [...]

Reinstall

* deplist package1 [package2] [...]

Show all dependencies for each package

* repolist [all | enabled | disabled]

List shows metadata information of yum warehouse

When the following parameters are not given, the default is enabled

* version [all | installed | available | group-* | nogroups* | grouplist | groupinfo]

* history [info | list | packages-list | packages-info | summary | addon-

Info | redo | undo | rollback | new | sync | stats]

View the history of yum transactions

When installing, upgrading, uninstalling, etc., yum will start the transaction

* load-transaction [txfile]

Restart a stopped transaction

* check

* help [command]

The yum command in CentOS7:

Administrative commands for the package group:

* groups [...]

Groups {install | remove | list | info | update}

Alias: groupinstall groupremove grouplist groupinfo groupudpate

* autoremove [package1] [...]

Automatic uninstall:

* repoinfo [all | enabled | disabled]

Commonly used third-party yum warehouse, Fedora EPEL

Go to the official site or mirror site, download the epel-release package, and use the rpm command to install to the corresponding Linux distribution, in the / etc/yum.repos.d directory, there will be an extra epel.repo file; update repolist to use

Linux virtual machine is required to connect to the Internet

Variables in yum's repo configuration file:

Releasever: the major version number of the current distribution of the operating system

For example, CentOS 7.2-1511, the major version number of its distribution is 7.

CentOS 6.7.The major version number of its distribution is 6.

$basearch: the type of infrastructure architecture used by the current computer

Such as: Intel compatible computer

I686

X86_64

Apple computer equipment

Ppc

Ppc64

$arch: platform architecture type

$YUM0~$YUM9:yum built-in 10 user-defined variables

Why use these variables?

The server (repository, repository) must be able to provide all the common distributions and platform-related packages; but the server cannot predict the specific version usage of any client.

So let the user choose:

.repo

Baseurl= https://mirrors.aliyun.com/epel/7/x86_64

Create a local yum repository:

Createrepo / PATH/TO/PACKAGES

Yum-config-manager-add-repo=URL

Compilation and installation of CentOS package Manager

Why compile and install the source code?

1. The corresponding package cannot be obtained directly

two。 The compiled binary does not contain the functions we need.

3. The compiled binary contains so many functions that we don't need

There are two kinds of source code packages:

Source code packaged directly by programmers: name-VERSION.tar.gz

Rpm source code package made by a third party using rpm tools: name-VERIONS-release.src.rpm

/ usr/src/name-VERIONS

The organization of the source code:

1. Multi-file composition

two。 There may be dependencies between the code in the file, even between the code snippets of different files

Project management tools for source code:

Project management tool: make

Java project management tool: maven

Source code-- > preprocessing (preprocessor, cpp)-- > compilation (compiler, gcc)-- > assembly (assembler)-- > link (linker)-- > binaries used to perform operations

For each source code file that needs to be compiled, there may be one or more files that he depends on or files that he depends on, and this information is saved to a configuration file; this configuration file tells us:

1. Which preprocessor is called and how to preprocess

two。 Which compiler to call and how to compile

3. The compilation order of files, etc.

Such files are usually makefile. Such makefile files do not exist in advance, nor do they treat the needs of all users equally.

Programmers can use special tools to analyze and extract the contents of all source code files and build makefile files. All makefile files have a unified template-Makefile.in.

Users can build makefile files according to their own needs, so as to realize the compilation process that can meet their own needs.

It is usually a script file called configure that helps users generate makefile files

Development tools:

Autoconf: help users generate configure scripts

Automake: help users generate template files for Makefile.in

Suggestion: any source code package, after decompression, will provide INSTALL or README documentation files. Before compiling, it is best to consult the contents of such files to determine the compilation process.

How to obtain the source code of an open source program:

1. Official self-built site:

Apache.org (ASF)

Mariadb.org

...

two。 The code's hosting site

SourceForge

Github

Code.google.com

Need to use compiler: gcc (GNU C Complier)

Installation of the development environment:

CentOS7:

1.yum groups install "development tools"

Yum groups install "Development Tools"

CentOS6:

Yum groupinstall "Development Tools"Server Platform Development"

"Desktop Platform Development"

2.yum install make automake autoconf gcc glibc qt ncurses

Generally speaking, there are three major steps in compiling an installer in a Linux system:

1.configure [options...]

2.make

3.make install

Step one:

. / configure [options...] [VAR=VALUE]...

Classification of common options:

1. Setting of the installation path:

-- prefix=/PATH/TO/SOMEWHERE

Specify the default program installation path, if this option is not given, the default path / usr/local/

-- sysconfdir=DIR

Specify the installation path for the configuration file

two。 System type:

3. Optional properties:

-- disable-FEATURE

Disable a feature or function during compilation

-- enable-FEATURE [= ARG]

Enable a feature or function during compilation

4. Optional package:

-- with-PACKAGE [= ARG]

During the compilation process, choose to use additional packages

-- without-PACKAGE

No additional packages are used during compilation

Step 2:

Make

Step 3:

Make install

Compilation closing work:

/ usr/local/apache2 will contain bin, sbin, etc, lib, lib64, include, man,...

1. Export binaries:

Create a file: / etc/profile.d/httpd.sh

Export PATH=/usr/local/apache2/bin:/usr/local/apache2/sbin:$PATH

two。 Export the library file:

Create a file: / etc/ld.so.conf.d/httpd.conf

/ usr/local/apache2/lib64

Ldconfig [- v]

3. Export the header file:

In the / usr/include directory, create a symbolic link to the header file for compiling the installed software:

Ln-sv / usr/local/apache2/include / usr/include/httpd

4. Export the help file:

CentOS6: / etc/man.config

Add a new line: MANPATH=/usr/local/apache2/man

CentOS7: / etc/man_db.conf

Add a new line: MANPATH_MAP=/usr/local/apache2/bin / usr/local/apache2/man

Example installation of httpd-2.4.18

1. The first step is to solve the development tools environment.

Yum groupinstall "Development tools"

two。 Use apr and apr-util tools for httpd-2.4.18 versions

Install apr tools

~] # tar xf apr-1.5.2.tar.bz2-C / usr/src

~] # cd / usr/src/apr-1.5.2

~] # / configure-- prefix=/usr/local/apr-1.5.2

~] # make

~] # make install

Install arp-util tools

~] # tar xf apr-util-1.5.4.tar.bz2-C / usr/src

~] # cd / usr/src/apr-util-1.5.4

] # / configure-- prefix=/usr/local/apr-util-1.5.4-- with-apr=/usr/local/apr-1.5.2

~] # make

~] # make install

3. Pcre-devel needs to be installed to meet the regular expression usage requirements of httpd

Yum-y install pcre-devel

Compile and install httpd:

~] # tar xf httpd-2.4.18.tar.gz-C / usr/src

~] # cd / usr/src/httpd-2.4.18

] # / configure-- with-apr=/usr/local/apr-1.5.2-- with-apr-util=/usr/local/apr-util-1.5.4

~] # make

~] # make install

Closing:

1. Export binaries:

Vim / etc/profile.d/httpd.sh

Export PATH=/usr/local/apache2/bin:$PATH

Source / etc/profile.d/httpd.sh

two。 Export the library file:

Vim / etc/ld.so.conf.d/httpd-x86_64.conf

/ usr/local/apache2/modules

Ldconfig-v

3. Export the header file:

Ln-sv / usr/local/apache2/include / usr/include/httpd

Sed command:

Last line command of vim (ex mode command):

Sed:Stream EDitor, the stream editor, edits one or more files in behavioral units; each time sed processes a line of content in a given file.

When sed processes text, the current line being processed is stored in a temporary buffer called "pattern space"; the current line is used to match the given PATTERN, and if it can be matched, command editing is used; if it does not match, the default output is output to standard output; and then proceed to the next line until the end of the file.

By default, all processing actions of sed do not affect the contents of the source file

We usually use the sed command to automatically edit one or more files to simplify the circular operation of the files.

Sed command:

Sed-stream editor for filtering and transforming text

Sed [OPTION]... 'script' [input-file]...

Common options:

-n,-- quiet,-- silent: lines that do not match patterns are not output to the screen by default

-e script,-- expression=script: multi-conditional editing

-f script-file,-- file=script-file: read the script from the specified script-file

Note: script-file is a path that can locate the script file, preferably an absolute path

-I [SUFFIX],-- in-place [= SUFFIX]: direct editing of the original file

-r,-- regexp-extended: enable sed to support extended regular expressions

Script:

'AddressesCommand'

Addresses (address delimiting):

1. Empty address: all lines of all files are processed

two。 Single address: sed processes the only line that can match that address

Num: indicates the line number

/ pattern/: can match all lines of the pattern

$: represents the last line of the document

3. Address range:

Addr1,addr2: all lines in the middle from addr1 to the end of addr2

For example: 2Jing 8

First~step: start with the line number marked by first and take the number represented by step as the step

For example: 1-2, 2-2

Addr1,+N: start with addr1, including the line of addr1, and continue to calculate N lines backwards

For example, 2 is equivalent to 2, 2, 2, 2, 7

Addr1,~N: start with addr1, including the line of addr1, and calculate the addr1*N line backward

For example, 2 ~ 5 is equivalent to 2 ~ 2 ~ 2 ~ 5

/ pattern1/,/pattern2/: is calculated from the first line matched by pattern1 to the end of the first line matched by / pattern2/

Command (processing command):

=: displays the line number of the row matched by the pattern

For example: sed-n'/ ^ $/ ='/ etc/grub2.cfg

A\ text: appends the contents of text to the lines that are matched by the pattern, and supports the use of line breaks to achieve multiple line appends.

I\ text: inserts the contents of text in front of the rows that are matched by the pattern, and supports the use of line breaks to achieve multi-line insertion.

C\ text: change the line matched by the pattern to the content of text, and support the use of\ nnewline, so as to change one line into multiple lines.

D: delete rows that are matched by patterns in the pattern space; such rows can no longer be stdout

Note: the-n option should not be used when using the d command

P: displays the rows in the pattern space that are matched by patterns

Note: when using the p command, it is usually accompanied by the-n option

W filepath: save the lines in the pattern space that are matched by the pattern to the filepath file

R filepath: appends the contents of the filepath file to the lines in the pattern space that are matched by the pattern

! Command: lines that are matched by patterns in pattern space do not execute the Command command; on the contrary, lines that are not matched by patterns execute the Command command

Find and replace. The delimiter can be changed at will, as long as it is the same.

S, s, s

S/pattern/text/ [control]

Swatches _ petchure _

Pattern: what you plan to find and replace

Text: the result to replace

Control: how to replace

G: replace all in the line

P: show the rows that have been replaced successfully

W filepath: save the replaced lines to the filepath file

Support for backward references:

S /\ (string\) / & /

S /\ (string\) /\ 1 /

Advanced editing commands:

H: overwrite the contents of the schema space to the holding space

H: append the contents of the schema space to the holding space

G: take the data from the hold space and overwrite it to the schema space

G: take out the content from the hold space and append it to the pattern space

X: interchange the content in the pattern space with the content in the hold space

N: read the next row of the matched row to overwrite the pattern space

N: append the next line of the matched row to the pattern space

D: delete rows in schema space

D: delete all rows in multiline mode space

Example:

Sed-n'nscape p 'FILE

Sed'1, Gentleman, FILE.

Sed'$! n 'FILE

Sed'$! d 'FILE

Sed 'G'FILE

Sed 'g' FILE

Sed'/ ^ $/ dscape G'FILE

Sed 'ntrandd'FILE

Sed-n'1g / n / 1g / h / n / n / 1 / g / h / n / 1 / n / n / 1 / 2 / 2 / 2 / 3 / 3 / 3 / 3 / 3 / 3 / 3 / 3 / 3 / 3 / 3 / 3 / 3 / 3 / 3 / 3 / 3 / 3 / 3 / 2 / 3 / 3 / 3 / 3 / 3 / 3 / 1 / 2 / 1 / 2 / 1 / 2 / 2 / 1 / 2 / 1 / 2 / 1 / 2 / 1 / 2 / 1 / 2 / 1 / 2 / 1 / 2 / 1 / 2 / 1 / 2 / 1 / 3 / 1 / 2 / 1 / 3 / 1 / 2 / 1 / 2 / 1 / 2 / 1 / 2 / 1 / 2 / n'1 / n'1 / 2 /

Grep: text search

Sed: text editing tool

Awk: file formatting tool

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.