Bug 9456: (follow-up) add tooltip
[koha.git] / INSTALL.debian
index c6e2003..4547974 100644 (file)
@@ -1,15 +1,14 @@
 =============================
-Installation Guide for Installing Koha on Debian Etch with MySQL 5
+Installation Guide for Installing Koha on Debian
 =============================
 
-Copyright (C) 2007, 2008 LibLime (http://liblime.com)
-
-Maintainer: Joshua Ferraro (jmf AT liblime DOT com)
+Some parts copyright 2010 Chris Nighswonger
+Some parts copyright 2011 MJ Ray and software.coop
 
 Feedback/bug reports: Koha Developer's List:
-http://lists.koha.org/mailman/listinfo/koha-devel
+http://lists.koha-community.org/cgi-bin/mailman/listinfo/koha-devel
 
-This document last modified: 30 April 2008
+This document last modified: 2012-March-20
 
 Installation Instructions
 =============================
@@ -19,141 +18,97 @@ as indicated or by running the command directly as root.
 
 1. Prepare System and Install Dependencies
 
-1.1 Install Debian Etch via Businesscard CD
-
-See http://www.debian.org/releases/etch/debian-installer/
-
-WARNING: use Perl 5.8 (a library Koha depends on,
-MARC::File::XML may not work with Perl 5.10, see:
-http://bugs.koha.org/cgi-bin/bugzilla/show_bug.cgi?id=2309),
+1.1 Install Debian via Business Card CD
 
-1.2 Set up your locale
+    See http://www.debian.org/CD/netinst/
 
-Your locale should be set to UTF-8, as should Apache2 and MySQL 5.
-This step is VERY IMPORTANT for a UNICODE compliant system. Please 
-read over the following document carefully:
+1.2 Set up Indexdata apt sources for Yaz and Zebra packages
 
-http://wiki.koha.org/doku.php?id=encodingscratchpad
+See the instructions from indexdata at
+http://ftp.indexdata.com/pub/debian/README
 
-IMPORTANT:  You _MUST_ follow all the steps outlined there for
-            Apache2, MySQL 5, etc. BEFORE you install Koha.
-
-1.3 Set up apt sources for Git, Yaz and Zebra packages
-
-Edit your /etc/apt/sources.list file and add the following:
+At the time of writing, you need to create
+/etc/apt/sources.list.d/indexdata.list containing the following:
 
     # Index Data
-    deb http://ftp.indexdata.dk/debian etch main
-    deb-src http://ftp.indexdata.dk/debian etch main
+    deb http://ftp.indexdata.dk/debian squeeze main
+    deb-src http://ftp.indexdata.dk/debian squeeze main
 
-    # Backports
-    deb http://www.backports.org/debian etch-backports main contrib non-free
+and add their archive signing key to your system:
 
-Backports packages are signed with a key you can install as follows:
+    $ wget -O- http://ftp.indexdata.dk/debian/indexdata.asc | sudo apt-key add -
 
+Finally, update your apt sources:
     $ sudo apt-get update
-    $ sudo apt-get install debian-backports-keyring
-
-The Index Data packages are signed with a key you can install as follows:
 
-    $ wget http://ftp.indexdata.dk/debian/indexdata.asc
-    $ sudo apt-key add indexdata.asc
+1.3 Get Koha
+Choose one of these options:
 
-Run the following command to update your system:
+1.3.1 Option A: Download Koha via Git (optional)
 
-    $ sudo apt-get update
-    $ sudo apt-get -t etch-backports install git-core git-email
-    $ sudo apt-get install yaz idzebra-2.0 idzebra-2.0-doc
-
-1.4 Get Koha
-
-1.4.1 Option A: Download Koha via Git (optional)
-
-    $ git clone git://git.koha.org/pub/scm/koha.git kohaclone
-    $ cd kohaclone
+    $ sudo apt-get install git-core git-email
+    $ git clone git://git.koha-community.org/koha.git koha
+    $ cd koha
     $ git checkout -b myinstall origin
 
 Note: for more information about Git, please see the Koha Git Usage Guide:
 
-    http://wiki.koha.org/doku.php?id=en:development:git_usage
+    http://wiki.koha-community.org/wiki/Version_Control_Using_Git
 
-1.4.2 Option B: Download Koha from http://download.koha.org
+1.3.2 Option B: Download Koha from http://download.koha-community.org
 
-    $ wget http://download.koha.org/koha-3.00.00-beta.tar.gz
-    ( Note: use the latest stable version)
+Find the latest Koha stable version on http://download.koha-community.org
+and download it to your server with something like:
 
-1.5 Install additional Debian dependencies
+    $ wget <URL found on download.koha-community.org>
+
+1.4 Install additional Debian dependencies
 
 IMPORTANT:  You should only use CPAN for Perl dependencies which are NOT
             available from the package maintainer. You have been warned!
 
-Using the debian.packages file included in the Koha source tree,
-run the following:
+Installing all necessary Debian packages, MySQL Server,
+Zebra and all Debian packaged Perl module required by Koha:
 
     $ sudo dpkg --set-selections < install_misc/debian.packages
+    $ sudo apt-get dselect-upgrade
 
-Now start dselect:
-
-    $ sudo dselect
-
-Choose [I]nstall and accept packages to be installed (hit return)
-
-(may take a while)
-
-Choose [C]onfigure, [R]emove and [Q]uit until dselect has completed.
-
-1.6 Install Perl dependencies that aren't packaged into Debian Etch
-sources
+1.5 Install Perl dependencies that aren't packaged into Debian
 
 Run the following command:
 
-    $ sudo cpan MARC::Record Class::Accessor MARC::Charset MARC::File::XML \
-                Net::Z3950::ZOOM HTML::Template::Pro MARC::Crosswalk::DublinCore \
-                PDF::Reuse PDF::Reuse::Barcode Data::ICal GD::Barcode::UPCE \
-                XML::RSS Algorithm::CheckDigits::M43_001 Biblio::EndnoteStyle POE Schedule::At
+    $ perl koha_perl_deps.pl -m -u
 
-There is a known but benign error in the test case for Barcode::Code128, 
-which is required by PDF::Reuse::Barcode.  If this module is not installed, 
-you can do a forced installation of Barcode::Code128.
+This will list whether any Perl dependencies are missing or need a
+later version and whether they are required.  You can then install
+them another way.  There may be packages on
+http://debian.koha-community.org or maybe you can build packages
+yourself or maybe you will need to install them locally with a command
+similar to
+
+    $ sudo cpan Graphics::Magick CHI CHI::Driver::Memcached
 
 Note: you may need to run CPAN initialization if you've not run cpan
-before:
-
-    /etc/perl/CPAN/Config.pm initialized.
-    
-    CPAN is the world-wide archive of perl resources. It consists of about
-    100 sites that all replicate the same contents all around the globe.
-    Many countries have at least one CPAN site already. The resources
-    found on CPAN are easily accessible with the CPAN.pm module. If you
-    want to use CPAN.pm, you have to configure it properly.
-    
-    If you do not want to enter a dialog now, you can answer 'no' to this
-    question and I'll try to autoconfigure. (Note: you can revisit this
-    dialog anytime later by typing 'o conf init' at the cpan prompt.)
-    
-    Are you ready for manual configuration? [yes]
+before.  See http://search.cpan.org/~andk/CPAN/lib/CPAN.pm#CONFIGURATION
 
 When the configuration is completed CPAN will install the Perl modules.
 
 2. Configuration of dependencies
 
-2.1 Update root MySQL password (if dselect didn't do it for you already)
-
-    $ sudo mysqladmin password <password>
-
-2.2 Create the Koha database
+2.1 Create the Koha database
 
-    Create the database and user with associated privileges:
+ Create the database and user with associated privileges (information inside <> brackets is
+ data you assign specifically to your installation. Do not include the <>'s when entering the commands):
 
     $ mysqladmin -uroot -p<password> create <kohadatabasename>
     $ mysql -uroot -p<password>
-    
+
     Welcome to the MySQL monitor.  Commands end with ; or \g.
     Your MySQL connection id is 22
-    Server version: 5.0.32-Debian_7etch3-log Debian etch distribution
-    
+    Server version: 5.0.51a-24 (Debian)
+
     Type 'help;' or '\h' for help. Type '\c' to clear the buffer.
-    
+
     mysql> grant all on <kohadatabasename>.* to '<kohadatabaseuser>'@'localhost' identified by '<kohadatabaseuserpassword>';
     Query OK, 0 rows affected (0.00 sec)
 
@@ -162,175 +117,170 @@ When the configuration is completed CPAN will install the Perl modules.
 
     mysql> quit
 
-2.3 Test your SAX Parser and correct where necessary
+You may want to document the database name, the database username, and the database password you just set.
+Step 3 will require them.
+
+2.2 Test your SAX Parser and correct where necessary
 
-    You must be sure you're using the XML::LibXML SAX parser, not Expat or PurePerl, both of which have outstanding bugs with pre-composed characters. You can test your SAX parser by running:
+You must be sure you're using the XML::LibXML SAX parser, not Expat or
+PurePerl, both of which have outstanding bugs with pre-composed characters.
+You can test your SAX parser by running:
 
     $ cd koha
-    $ misc/sax_parser_print.pl 
+    $ misc/sax_parser_print.pl
 
-    You should see something like::
+You should see something like:
 
     XML::LibXML::SAX::Parser=HASH(0x81fe220)
 
-    If you're using PurePerl or Expat, you'll need to edit your
-    ini file, typically located at:
-
-    /etc/perl/XML/SAX/ParserDetails.ini
-
-2.4 Install DBD::mysql Perl module
-
-In order to handle UTF-8 correctly, Koha requires at least version 4.004
-of the DBD::mysql Perl module. However, Debian Etch has a stable package
-only for version 3.0008, so it is necessary to install the module  from CPAN.
-DBD::mysql's test suite needs to use a MySQL 'test' DB which doesn't exist 
-anymore. So there are two options to install DBD::mysql: 
+If you are configured to use PurePerl or Expat, the script will say you have a problem and
+you'll need to edit your ini file to correct it.
 
-  (1) install without test suite,
-  (2) install with test suite requiring a test MySQL DB creation.
+The file is typically located at:
 
-2.4.1 Install without test suite
-
-Force install DBD::mysql:
+    /etc/perl/XML/SAX/ParserDetails.ini
 
-    $ sudo cpan
-    cpan> force install DBD::mysql
+The correct parser entry will need to be moved to the bottom of the file.
+The following is the entry you are looking for:
 
-2.4.2 Create test database in order to install DBD::mysql
+    [XML::LibXML::SAX::Parser]
+       http://xml.org/sax/features/namespaces = 1
 
-Because of DBD::mysql's test suite, it is necessary to temporarily create a
-test database and user:
+2.3 Create your Koha system user (if you haven't created one already)
 
-    $ mysql -uroot -p<password>
+    $ sudo adduser koha
 
-    Create the database and user with associated privileges:
+You can substitute any username for "koha," just be sure to write it down, you will need to know what it is in step 3.
 
-    Welcome to the MySQL monitor.  Commands end with ; or \g.
-    Your MySQL connection id is 22
-    Server version: 5.0.32-Debian_7etch3-log Debian etch distribution
+3. Run the Koha installer
 
-    Type 'help;' or '\h' for help. Type '\c' to clear the buffer.
+    $ perl Makefile.PL
+      ( answer questions )
+    $ make
+    $ make test
+    $ sudo make install
 
-    mysql> create database test;
-    Query OK, 1 row affected (0.00 sec)
+3.1 Export the environment variables
+Make install will tell you what the proper path for these environment variables is.
+In case you missed it at the end of make install, here are some examples:
 
-    mysql> grant all on test.* to 'test'@'localhost' identified by 'test';
-    Query OK, 0 rows affected (0.00 sec)
-    (test database, user, and password can be different if need be)
+Example (for standard install):
+    $ export KOHA_CONF=/etc/koha/koha-conf.xml
+    $ export PERL5LIB=/usr/share/koha/lib
 
-    mysql> flush privileges;
-    Query OK, 0 rows affected (0.00 sec)
+Example (for dev install. These paths are set during Makefile.PL.):
+    $ export KOHA_CONF=/path/to/your/koha-conf.xml
+    $ export PERL5LIB=/path/to/koha/lib
 
-    mysql> quit
+4. Configure and start Apache
+    $ sudo ln -s /etc/koha/koha-httpd.conf /etc/apache2/sites-available/koha
+(note that the path to koha-httpd.conf may be different if you chose a dev install)
 
-Next install DBD::mysql:
+Add the following lines to /etc/apache2/ports.conf:
 
-    $ sudo cpan
+    Listen 80
+    Listen 8080
 
-    cpan> o conf makepl_arg
-    (get current value of this CPAN parameter)
+Run the following commands:
 
-    cpan> o conf makepl_arg "--testdb=test --testuser=test --testpass=test"
+    $ sudo a2enmod rewrite deflate
+    $ sudo a2ensite koha
+    $ sudo apache2ctl restart
 
-    cpan> install DBD::mysql
+Note: you may still see the usual Apache default site if your VirtualHost
+      configuration isn't correct.  If Koha is the only web application running on the server,
+      the command "sudo a2dissite default" may be a quick fix. For servers running other sites
+      alongside Koha, see the Apache HTTPD manual section on virtual hosts for full
+      instructions (http://httpd.apache.org/docs/2.2/vhosts/).
 
-    cpan> o conf makepl_arg ''
+5. Run the Web Installer, populate the database, initial configuration of settings
 
-    OR 
+Point your browser to http://<servername>:8080/
 
-    cpan> o conf makepl_arg '<old setting>'
+Note: <servername> is (usually) the IP of your server, or localhost (if you are connecting
+      from the same machine Koha is installed on. You can verify the location by checking
+      the VirtualHost settings for both the opac and intranet virtual hosts in koha-httpd.conf
 
-    (restore this setting so as to not interfere with future CPAN installs).
+Koha will redirect you to the Web Installer where you can continue the setup. You will be prompted to enter in your DATABASE username and password. It is through this interface
+that
+you can install the sample data for libraries, patrons, and much more.
 
+Be aware that removing sample data at a later time (when you may want to take the
+server into production) may not be easy, and if your intent is to take this
+install from testing to production, go with minimal sample data (no patrons or bibliographic records).
 
-Finally, remove the test database:
+6. Configure and start Zebra
 
-    $ mysql -uroot -p<password>
+Note: it's recommended that you daemonize the Zebra process and add it to your
+      startup profile. For a non-production test/development installation, running
+      Zebra from the command line can be useful. Pick one of the options below (or roll your own).
 
-    mysql> drop database test;
-    Query OK, 1 row affected (0.00 sec)
+Note: it's also recommended that you create a Koha system user, which you will
+      have specified during the install process. Alternatively, Zebra can be
+      configured to run as the root user.
 
-    mysql> exit
-    Bye
+Option 1: run the Zebra processes from the command line (manual indexing):
 
-3. Run the Koha installer
+6.1.1 Zebra Search Server
 
-    $ perl Makefile.PL
-    ( answer questions )
-    $ make
-    $ make test
-    $ sudo make install
+This process send responses to search requests sent by Koha or
+Z39.50/SRU/SRW clients.
 
-4. Configure and start Apache
-    $ sudo ln -s /etc/koha/koha-httpd.conf /etc/apache2/sites-available/koha
-    (note that the path to koha-httpd.conf may be different depending on your
-    installation choices)
+    $ sudo -u <kohauser> zebrasrv -f /etc/koha/koha-conf.xml
+    (note that the path to koha-conf.xml may be different if you chose a dev install, and that
+    <kohauser> refers to the user you set up in step 2.3)
 
-Add the following lines to /etc/apache2/ports.conf:
+Note: the user you run Zebra as will be the only user with write permission
+      on the Zebra index; in development mode, you may wish to use your
+      system user.
 
-    Listen 80
-    Listen 8080
 
-Run the following commands:
+6.1.2 Zebra Indexer
 
-    $ sudo a2enmod rewrite
-    $ sudo a2ensite koha
-    $ sudo apache2ctl restart
+Added/updated/deleted records in Koha MySQL database must be indexed
+into Zebra. A specific script must be launched each time a bibliographic
+or an authority record is edited.
 
-5. Configure and start Zebra
+    $ sudo -u <kohauser> misc/migration_tools/rebuild_zebra.pl -z -b -a
 
-Note: it's recommended that you daemonize the Zebra process and add it to your
-startup profile. For a non-production test/development installation, running 
-Zebra from the command line can be useful. Pick from the two available options
-below, or roll your own :-)
+Note: This script should be run as the kohauser (the default is 'koha', but
+this is the user you set up in step 2.3).
 
-Note: it's also recommended that you create a Koha system user, which you will
-have specified during the install process. Alternatively, Zebra can be
-configured to run as the root user.
+Option 2: run the Zebra process as a daemon (automatic indexing):
 
-Option 1: run the Zebra processes from the command line:
+Note: References to <script_dir> refer to the directory where
+      Koha's command-line scripts are installed, the path
+      is /usr/share/koha/bin/ by default in a standard install.
 
-    Zebra Server:
-    $ sudo -u ${KOHA_USER} zebrasrv -f /etc/koha/koha-conf.xml
-    (note that the path to koha-conf.xml may be different depending on your
-    installation choices)
+6.2.1 Zebra Search Server
 
-    Note: the user you run Zebra as will be the only user with write permission
-    on the Zebra index; in development mode, you may wish to use your system
-    user.
+    $ sudo ln -s <script_dir>/koha-zebra-ctl.sh  /etc/init.d/koha-zebra-daemon
+    (Note: <script_dir> is /usr/share/koha/bin/ by default in a standard install)
+    $ sudo update-rc.d koha-zebra-daemon defaults
 
-    Zebraqueue Daemon:
-    $ sudo -u ${KOHA_USER} misc/bin/zebraqueue_daemon.pl
+    $ sudo <script_dir>/koha-zebra-ctl.sh start
+    (Note: <script_dir> is /usr/share/koha/bin/ by default in a standard install)
 
-    Note: if you are running in this mode, you may wish to defer starting the
-    zebraqueue until you've run the web installer (see below).
 
-Option 2: run the Zebra processes as daemons, and add to startup process:
+6.2.2 Zebra Indexer
 
-Note that references to $SCRIPT_DIR refer to the directory where
-Koha's command-line scripts are installed, e.g., /usr/share/koha/bin.
+Add an entry in Koha user crontab to process scheduled added/updated/deleted records
+indexing by Zebra. <script_dir>cronjobs/crontab.example contains examples for these cron jobs (and many more).
 
-    Zebra Server:
-    $ sudo ln -s ${SCRIPT_DIR}/koha-zebra-ctl.sh  /etc/init.d/koha-zebra-daemon
-    (Note: ${SCRIPT_DIR} is /usr/share/koha/bin/ by default in a standard install)
-    $ sudo update-rc.d koha-zebra-daemon defaults
-    ( Note: see man chkconfig(8) on other distros )
-    
-    $ sudo ${SCRIPT_DIR}/koha-zebra-ctl.sh start
+NOTE: The cronjobs should be setup under the kohauser (the default is 'koha', but
+this is the user you set up in step 2.3).
 
-    Zebraqueue Daemon:
-    $ sudo ln -s ${SCRIPT_DIR}/koha-zebraqueue-ctl.sh  /etc/init.d/koha-zebraqueue-daemon
-    $ sudo update-rc.d koha-zebraqueue-daemon defaults 
-    ( Note: see man chkconfig(8) on other distros )
+Edit the crontab for the koha user by running
+    $ sudo -u <kohauser> crontab -e
 
-    $ sudo ${SCRIPT_DIR}/koha-zebraqueue-ctl.sh start
+For Zebra indexing, you are looking for the example that begins with
 
-6. Run the Web Installer, populate the database, initial configuration of settings
+    # ZEBRA INDEX UPDATES with -z option, incremental index updates throughout the day
+    # for both authorities and bibs
 
-    Point your browser to http://<servername>:8080/
+It may be easiest to copy/paste the example into your own crontab and modify as necessary.
 
-    It should redirect you to the Web Installer where you can continue the setup.
-    You can install the sample data for libraries, patrons, etc. via the Web Installer
+You can also configure zebra-indexing as an background daemon, see http://wiki.koha-community.org/wiki/Background_indexing_with_Zebra
 
 7. What next?
 
@@ -342,29 +292,64 @@ Koha's command-line scripts are installed, e.g., /usr/share/koha/bin.
       (note: use the correct path to your koha-conf.xml)
 
 7.1 Import:
-    $ misc/migration-tools/bulkmarcimport.pl -file /path/to/marc.iso2709
+
+Bibliographic data in MARC21 format
+
+    $ misc/migration_tools/bulkmarcimport.pl -file /path/to/marc.iso2709
+
+Authority data in MARC21 format
+
+    $ misc/migration_tools/bulkmarcimport.pl -a -file /path/to/auth.iso2709
 
 7.2 Fast Index:
-    $ misc/migration-tools/rebuild_zebra.pl -b -w
+    $ misc/migration_tools/rebuild_zebra.pl -b -w
+
+    Once the indexing has completed, you will be able to search for records in your system.
+    NOTE: This script should be run as the kohauser (the default is 'koha', this is the user we set up in step 2.3).
+
+7.3 Schedule regular index updates with cron ,or configure zebra indexing as a background daemon
 
-    Once the indexing has completed, you will be able to search for records in your
-    system. NOTE: if you don't run the Fast Index utility, and you have the zebraqueue
-    daemon running, the indexin process will happen in the background, but it will be
-    orders of magnitude slower.
+    You need to run rebuild_zebra.pl -b -a -z as a regular cron job in orde to pick up new bibs
+    and items as you add them. Check misc/cronjobs/crontab.example for usage examples. See 7.0 above.
+    NOTE: This job should be setup under the kohauser (the default is 'koha', this is the user we set up in step 2.3).
 
-    To enable the SRU servers, you'll need to edit your koha-conf.xml and change the
-    <listen> options to listen on a TCP port; then restart the zebra daemon.
+    To setup indexing in background see 6.2.2
+
+7.4 To enable public Z39.50/SRU servers, you'll need to edit your koha-conf.xml and
+    change the <listen> options to listen on a TCP port; then restart the zebra daemon.
 
 UPGRADE
 =======
-If you are upgrading from a previous installation of Koha 3, you can
+If you are running in another language other than english, please
+switch to english before doing the upgrade, the templating system has
+changed and the templates will need to be regenerated.
+Once you have upgraded, please regenerate your templates in your
+chosen languages.
+
+If you are upgrading from a previous installation of Koha 3.x, you can
 use the following:
 
+ ./koha_perl_deps.pl -u -m # to identify new Perl dependencies
+
+Install any missing modules
+IMPORTANT: Koha 3.6.x uses Template::Toolkit, this must be installed
+before the webinstaller can run
+
+ sudo apt-get install libtemplate-perl
+
  perl Makefile.PL --prev-install-log /path/to/koha-install-log
  make
  make test
  sudo make upgrade
 
+Koha 3.4.x or later no longer stores items in biblio records so
+if you upgrading from an old version as part of the
+upgrade you will need to do the following two steps, they can take a
+long time (several hours) to complete for large databases
+
+ misc/maintenance/remove_items_from_biblioitems.pl --run
+ misc/migration_tools/rebuild_zebra.pl -b -r
+
 Uninstall Instructions
 =============================
 1) Stop Services:
@@ -373,11 +358,9 @@ Uninstall Instructions
    $ sudo apache2ctl restart
 
    $ sudo update-rc.d koha-zebra-daemon remove
-   $ sudo update-rc.d koha-zebraqueue-daemon remove
    $ sudo rm /etc/init.d/koha-zebra-daemon
-   $ sudo rm /etc/init.d/koha-zebraqueue-daemon
 
-2) Remove Koha Databases
+2) Remove Database and Indexes
 
    # MySQL
    $ mysql -u<kohauser> -p<kohapassword>
@@ -388,14 +371,12 @@ Uninstall Instructions
    $ zebraidx -c <prefix>/etc/zebradb/zebra-authorities.cfg -g iso2709 -d authorities init
 
 3) Remove Koha Install Directories and Configuration Files
+   Don't forget about any crontab entries
 
 Tested on the following operating environments
 ==============================================
-- Debian Etch 4.0
-
-Installer Bug reports
-=====================
-Please send any installer bug reports to jmf AT liblime DOT com
+- Debian Lenny 5.0
+- Debian Squeeze 6.0
 
 Other Notes
 =====================
@@ -412,3 +393,4 @@ A PARTICULAR PURPOSE.  See the GNU General Public License for more details.
 You should have received a copy of the GNU General Public License along with
 Koha; if not, write to the Free Software Foundation, Inc., 59 Temple Place,
 Suite 330, Boston, MA  02111-1307 USA
+