Dbconfig common настроить или нет

Как пользоваться phpMyAdmin на Ubuntu

Всем привет! В одной из недавних статей я рассказал об установке LAMP на Ubuntu — локального сервера для Linux. К сожалению, средство администрирования баз данных phpMyAdmin не входит в набор инструментов LAMP, поэтому устанавливается отдельно. Из этой статьи вы узнаете, как пользоваться phpMyAdmin.

Установка phpMyAdmin на Ubuntu

Очень часто phpMyAdmin (PMA) встречается на платных хостингах и используется для создания и администрирования баз данных (БД) сайтов. Устанавливать ее на платный хостинг не надо — там уже все сделано до вас. Чаще всего вы будете сталкиваться с установкой PMA на свой компьютер для администрирования баз данных локального сервера. В качестве примера я рассмотрю установку phpMyAdmin на Ubuntu — один из самых популярных дистрибутивов Linux.

Откройте терминал и введите следующую команду:

После этого введите пароль администратора и согласитесь с продолжением установки.

Выберете веб-сервер, который автоматически будет настроен для запуска PMA. В нашем случае это apache2.

Соглашайтесь настроить БД с помощью dbconfig-common.

Введите пароль администратора MySQL.

Задайте пароль для регистрации PMA на сервере баз данных и далее подтвердить его. После окончания установки, phpMyAdmin будет доступна по адресу http://localhost/phpmyadmin .

phpMyAdmin не работает

Если перейдя по этому адрес, вы увидите страницу с ошибкой, то можно сделать следующее.

введите в терминале команду:

Пользователь будет root , а пароль вы задали при установке .

Создание новой базы данных

Пройдя аутентификацию, мы попадаем на стартовую страничку PMA, на которой можно выбрать язык интерфейса, если вы ранее этого еще не сделали, посмотреть информацию о системе управления базами данных MySQL и используемом веб-сервере. В левом же столбике представлены все имеющиеся БД.

Что в первую очередь пригодится? Конечно, умение создавать БД для будущего сайта. Для этого перейдите на вкладку Базы данных .

Достаточно ввести название новой БД и кликнуть на кнопку «Создать» — все остальное phpMyAdmin сделает автоматически.

Появится сообщение об успешном создании БД и она добавится в список ниже.

Добавление нового пользователя БД

Создав новую БД, необходимо еще и добавить пользователя, который бы мог ей пользоваться.

Для этого кликните на надписи «Проверить привилегии». Вас перебросит на страницу «Пользователи с правами доступа к БД», на которой следует кликнуть на «Добавить нового пользователя».

В разделе «Информация учетной записи» укажите имя пользователя (английскими буквами), хост (на локальном сервере — localhost) и пароль.

Чуть ниже необходимо задать привилегии для нового пользователя. На локальном сервере можно не задумываясь отметить все пункты галочками. После этого следует нажать на кнопку «ОК» в самом низу и новый пользователь будет создан.

Импорт и экспорт баз данных

Для создания бэкапа (резервной копии) базы данных служит раздел «Экспорт».

В случае «быстрого» способа экспорта будут созданы бэкапы сразу всех доступных баз данных в формате, выбранном ниже.

Если необходимо создать резервную копию какой-то одной БД, то следует воспользоваться «обычным способом». В этом случае выберете нужную БД и в самом конце странице нажмите ОК.

В случае, если с базой данных что-то случилось и необходимо воспользоваться ранее созданной ее резервной копией необходимо:

1. Удалить поврежденную БД. Для этого перейдите на закладку «Базы данных», отметьте поврежденную БД и нажмите «Удалить»:
   
2. Загрузить (импортировать) рабочую версию БД. На вкладке «Импорт» нажмите кнопку Browse. (обзор), найдите файл бэкапа на своем компьютере и нажмите «ОК» в конце страницы.

На этом все. Как-то мало текста, но много картинок в статье получилось. Наверное, так и должно быть, ведь работать с phpMyAdmin не так сложно — главное знать, где и что находится, а это лучше всего показать на картинках.

Подписывайтесь на ленту новостей блога, чтобы всегда быть в курсе новых статей. Берегите себя!

Тогда ты просто обязан уметь работать с основными инструментами веб-мастера!

Где купить хостинг и домен? Как редактировать картинки? Как зайти на сервер, загрузить на ftp файлы и поменять права доступа? Как сменить кодировку? Обо всем этом и многом другом ты можешь узнать из бесплатного видеокурса «Инструменты Веб-мастера».

Лучший способ выразить благодарность автору — поделиться с друзьями!

Узнавайте о появлении нового материала первым! Подпишитесь на обновления по email:

Источник

dbconfig-common documentation
Chapter 3 — Using dbconfig-common in your packages

3.1 Quick and dirty: what to do

There are three things you will have to do as a package maintainer if you want to use dbconfig-common : provide the database code/scripts to setup the data base, source the maintainer script libraries and launch dbconfig-common . dbconfig-common will take care of everything else, including all debconf related questions, database/database-user creation, upgrade/remove/purge logic, etc. After all, the goal of dbconfig-common is to make life easier for both the local admin and the package maintainer 🙂

3.1.1 Update package dependencies

If your package just supports a single database type supported by dbconfig-common, your package needs to depend on the matching dbconfig- package as well as have the dbconfig-no-thanks as an alternative to that. E.g. packages that need a PostgreSQL database must have the following in their Depends field:

Packages that support multiple database types (see more about that in the Packages that support multiple types of databases, Section 3.2.4) must alternatively depend on all the matching dbconfig- packages as well as have dbconfig-no-thanks as an alternative to that. E.g. packages that need a PostgreSQL or SQLite3 database must have the following in their Depends field:

3.1.2 Putting hooks into the maintainer scripts

In the config , postinst , prerm , and postrm scripts for your package, you will need to source the libraries which perform most of the work for you (you do not need to do so in your preinst script). If you are not currently using debconf in your package, you will be now, and the debconf libraries need to be sourced first. You will need to use dh_installdebconf or otherwise install your config script into your deb file if you’re not already doing so. For example, here’s what it might look like in a config script for an imaginary foo-mysql package:

dbc_go is a function defined in every maintainer script hook to execute the appropriate code based on which maintainer script is being run. Note that it is passed two arguments. foo-mysql , the name of the package (there’s sadly no clean way to figure this out automatically), and $@ (the arguments which were passed to the maintainer script).

NOTE: you do not need to conditionally test for the existance of the shell library in the postinst and prerm scripts, but to stay compliant with Policy section 7.2 you do need to do so at least in your config and postrm scripts.

Note that if your package does not use debconf, you will need to explicitly install the config script in your package. The easiest way to do so is to call dh_installdebconf from debian/rules .

3.1.3 Supplying the data/code for your database

There are three locations in which you can place code for installing the databases of your package:

where PACKAGE is the name of the package, DBTYPE is the type of data (mysql, pgsql, etc). The full location should be a file containing the proper data.

The first location is for the majority of situations, in which the database can be constructed from it’s native language (SQL for MySQL/PostgreSQL, for example). The data will be fed to the underlying database using the credentials of the database user. The second location is like the first location, but will be run using the credentials of the database administrator. Warning: use of this second location should only be done when there are excerpts of database code that must be run as the database administrator (such as some language constructs in postgresql) and should otherwise be avoided. The third location is for databases that require a more robust solution, in which executable programs (shell/perl/python scripts, or anything else) can be placed.

This code will only be executed on new installs and reconfiguration of failed installs. In the case of SQL databases, in the data directory you would find the simple create and insert statements needed to create tables and populate the database. You do not need to create the underlying database, only populate it. The scripts directory contains shell/perl/python/whatever scripts, which are passed the same arguments as dbc_go . If you need database connection information (username, password, etc) in your scripts, you can source the /bin/sh format package config file, or you can instruct dbconfig-common to generate one in your programming language of choice (see the advanced tips section).

If files exist in both data and scripts, they will both be executed in an unspecified order.

That’s it! The rest of what needs to be done is handled by dbconfig-common , which should keep all the work (and bugs) in one place. Happy packaging! Of course, it’s recommended you take a quick look through the rest of the document, just to get an idea of other things that dbconfig-common can do for you in case you have special needs.

3.2 Advanced usage.

3.2.1 Generating custom configuration files with database information

Your database application will probably require a username and password in order to function. Every package that uses dbconfig-common already has a /bin/sh includable format config file, but it may be more convenient to have something in the native language of the package. For example, packaging a php/MySQL web app would be a lot easier if there were already a file existing with all the information in php includable format.

Using dbconfig-common , you can do this with little effort. In your postinst script, define the variable dbc_generate_include to a value that follows the form format:location where format is one of the supported output formats of dbconfig-generate-include (list them with -h) and location is the absolute location where you want your config file to go. There are also some extra variables dbc_generate_include_owner , dbc_generate_include_perms , and dbc_generate_include_args which do what you would expect them to. Note: you will be responsible for removing this file in your postrm script. Note2: the ownership and permissions will only be used when creating the file, so don’t rely on this feature if your packages wants to change existing ownership and/or permissions. When your scripts are run, this environment variable will be exported to your scripts, as well as a variable dbc_config_include which has the same value, but with the leading format: stripped away for convenience. NOTE if you use this feature, you should also ensure that the generated file is properly removed in the postrm script. dbconfig-common can not handle this itself, unfortunately, because it may be possible that it is purged before your package is purged. therefore, you should do the following in your postrm script:

3.2.2 Importing dbconfig-common into an existing package

If your package is already part of debian, dbconfig-common provides some support to load pre-existing settings from a specified config by setting two variables: dbc_first_version and dbc_load_include .

dbc_load_include should be specified in the config script and be of the format format:inputfile . format is one of the languages understood by dbconfig-load-include , and inputfile is either the config file in format language, or a script file in format language that otherwise determines the values and sets them.

dbc_first_version should be specified in both the config and postinst scripts, and should contain the first version in which dbconfig-common was introduced. When the package is installed, if it is being upgraded from a version less than this value it will attempt to bootstrap itself with the values.

3.2.3 Database changes in new versions of your package

Occasionally, the upstream authors will modify the underlying databases between versions of their software. For example, in MySQL applications column names may change, move to new tables, or the data itself may need to be modified in newer upstream versions of a package.

In order to cope with this, a second set of file locations exists for providing packagers ways to modify the databases during package upgrades:

where VERSION is the version at which the upgrade should be applied, and the respective path contains the upgrade code/data. When a package upgrade occurs, all instances of VERSION which are newer than the previously installed version will be applied, in order. There is also an automatically included set of safeguards and behavior provided by dbconfig-common , so as the packager you shouldn’t need to worry about most of the error-handling.

As with installation, scripts will be passed the same cmdline arguments as were passed to dbc_go.

3.2.4 Packages that support multiple types of databases

Sometimes, a particular package may support multiple database types. This is common with perl or php based web applications, which frequently use some form of database abstraction layer (pear DB for php, the DBD family for perl).

dbconfig-common provides support for such applications in a relatively straightforward fashion, allowing the local admin to select which database type to use when configuring a database for a package

To take advantage of this feature, you will want to use the «generic» maintainer script hooks, and additionally hint the debconf config script with the types of databases your package supports. For example, the postinst script would now look like this:

Unfortunately, specifying the proper dependencies on the right dbconfig- packages, as discussed earlier, is not enough for dbconfig-common to figure out which database types your package supports. Therefore you need to let dbconfig-common know via the config script. It needs to contain an additional variable called «dbc_dbtypes», which is a comma-separated list of supported database types:

3.2.5 Packages that connect to but should not create databases (read-only frontends)

Some packages provide multiple frontend packages to a single backend package. Furthermore, sometimes these frontend packages are installed on a seperate system from the actual database application, and should not manage the databases on their own.

For example, if the frontend were to be installed on multiple servers (perhaps load balancing or similar), it would not be wise to attempt to install/upgrade the database on each client. Instead, it would be wiser to simply prompt for the information and leave the database management to the single central package.

If the above scenario matches one of your packages, there are a seperate set of maintainer hooks for you to use. For example, frontend.config or frontend.config.mysql . Using these hooks, dbconfig-common will know enough to not take any actions apart from prompting the local administrator for the pertinent information.

3.2.6 Packages that require extra logic during removal

Sometimes, it may be that your install sql/scripts perform operations that aren’t automatically undone by package removal. For example, if your package gives extra grants to a user (such as triggers) it’s possible that grants will not automatically be revoked, which could cause problems for later installations as well as potential security concerns. For this and any other use you may need it for, you can place files in the following locations for «removal» maintainer code:

This works just like the install/upgrade code, only it always runs as the dbadmin. This code is run by default, unless the local admin opts out of deconfiguration assistance (note that this is seperate from database purging, which does not happen by default). Note that if you need to perform template substitution, you should set dbc_sql_substitutions to «yes» in your prerm maintainer script as well.

3.2.7 Hinting defaults and advanced control of configuration/installation

dbconfig-common has a set of pre-defined default values for most of the questions with which it prompts the user, most of which are variations on the name of the package. However, as a packager you can override some these values and set defaults that you feel are more appropriate, as well as otherwise modify the behavior of some parts of dbconfig-common .

The following table lists the variables you can hint in your config script, as well as some other variables you can use to have a finer level of control over dbconfig-common . You must use these variables exactly (and only) where directed in this table.

dbc_dbuser (used in: config )

Name to use when connecting to database. (defaults to: package name)

dbc_basepath (used in: config )

Database storage directory for local (filesystem) based database types. Not applicable for RDBMs like MySQL and postgres. (defaults to: /var/lib/dbconfig-common)

dbc_dbname (used in: config )

Name of database resource to which to connect. (defaults to: package name)

dbc_dbtypes (used in: config )

Database types supported by the package, in order of maintainers’ preference. (defaults to: empty)

dbc_dbfile_owner (used in: postinst )

Set the owner:group for the generated database file. This option is only valid for databases like SQLite that use a single file for storage and is not prompted via debconf. (defaults to: root:root)

dbc_dbfile_perms (used in: postinst )

Set the permissions for the generated database file. This option is only valid for databases like SQLite that use a single file for storage and is not prompted via debconf. (defaults to: 0640)

dbc_generate_include (used in: postinst )

format:outputfile pair for an extra config to be generated by dbconfig-generate-include . (defaults to: empty)

dbc_generate_include_owner (used in: postinst )

Set the owner:group of include files generated by dbconfig-generate-include . (defaults to: empty)

dbc_generate_include_perms (used in: postinst )

Set the permissions of include files generated by dbconfig-generate-include . (defaults to: empty)

dbc_generate_include_args (used in: postinst )

Arguments passed directly to dbconfig-generate-include . (defaults to: empty)

dbc_dgi_on_manual (used in: postinst )

Control whether config files should be generated by dbconfig-generate-include when the admin opts for manual installation. (defaults to: true)

dbc_first_version (used in: config , postinst )

The first version in which dbconfig-common was introduced in the package. (defaults to: empty)

dbc_load_include (used in: config )

format:includefile pair for a config to be read in by dbconfig-load-include . (defaults to: empty)

dbc_load_include_args (used in: config )

Arguments passed directly to dbconfig-load-include . (defaults to: empty)

dbc_pgsql_createdb_encoding (used in: postinst )

Specifies encoding for created postgres databases. (defaults to: empty/system default)

dbc_mysql_createdb_encoding (used in: postinst )

Specifies encoding for created MySQL databases. (defaults to: empty/system default)

dbc_sql_substitutions (used in: postinst , sometimes prerm )

If nonempty, specifies that sql files should be piped through a template substitution filter ( dbconfig-generate-include -f template ) before being executed. (defaults to: empty)

dbc_authmethod_user (used in config )

If set to «ident», dbconfig-common will set the default postgres authentication method for the package’s database user to «ident». (defaults to: empty)

dbc_config_allow_backup (used in config )

If nonempty, dbconfig-common will allow its state machine to backup past the beginning, such that packages allowing backup in there own config script to work transparently.

dbc_prio_low / dbc_prio_medium / dbc_prio_high / dbc_prio_critical (used everywhere)

If nonempty, dbconfig-common will use the values of these variables to set the priority of its debconf questions. You can use this if you think that the default levels of dbconfig-common are not well defined for your package. Be comforted thou that the priority will be raised automatically on errors and that all error handling allows the sysadmin to be asked the questions again.

dbc_authplugin (used in config )

This option defines which MySQL authentication plugin should be set to default whenever dbconfig-common creates a user in the database on behalf of the package using it. Choices are:

default: use the default auth plugin for installed MySQL server.

mysql_native_password: no MySQL authentication plugin is used.

sha256_password: more secure password encryption than native.

caching_sha2_password: in-memory authentication cache.

dbc_authplugin should be set in .config file, right before calling dbc_go function. It can also be configured by user during package installation, or during dpkg-reconfigure, if debconf questions priority is set to low.

Be careful when changing this. There are some differences in between MySQL and MariaDB and some options might not be available on one or another. As up to MariaDB 10.4, its default authentication plugin is «mysql_native_password» while, for MySQL 8.0, the new default authentication plugin is the «caching_sha2_password».

3.2.8 Debugging problems with dbconfig-common

In the event that your package is having trouble working with dbconfig-common , the first thing you should try is to export and set the shell variable dbc_debug to a nonempty value before installing your package. This will provide a slightly larger amount of information about what’s going on.

In the event that this does not provide enough information, the next thing to do will provide much, much, more information; enough that you will probably want to redirect stderr into a temporary output file. In the file /usr/share/dbconfig-common/dpkg/common , uncomment the set -x line near the top of the file. This will show you all the shell commands and logic as they are executed. If you have a good idea of where the problem is occurring, you can also insert your own set -x lines elsewhere followed by set +x lines to reduce the amount of input.

Источник