Installing Evergreen Trunk on Redhat

Obviously this document is a little dated now anyway, and the reader should know that as of May 2012 RHEL and CentOS targets in the prerequisites installer "Makefile.install" have been removed. It's not that we think it's impossible to get Evergreen working on these systems (we know it isn't, at least in the former case), but that we haven't had anyone share with the community up-to-date working Redhat configs or volunteer to test them. Start here if you want to resurrect what we had from our version control system to try to get it working again: http://git.evergreen-ils.org/?p=working/Evergreen.git;a=shortlog;h=refs/heads/user/dbs/update_makefile_prereqs

The following steps have been tested on Redhat x86 (32-bit) and x86-64 (64-bit) architectures.

In the following instructions, you are asked to perform certain steps as either the root user, the opensrf user, or the postgres user. To become the root user, issue the su - command and enter the password of the root user.

To switch from the root user to a different user, issue the su - <username> command; for example, su - opensrf. Once you have become a non-root user, to become the root user again simply issue the exit command.

Install the latest version of OpenSRF trunk. Follow the steps and run the test to ensure that OpenSRF is properly installed before continuing with any further Evergreen installation steps. Evergreen is an application that has been built on top of the Open Service Request Framework (OpenSRF), so if OpenSRF doesn't work, Evergreen isn't going to work.

Download and build Evergreen:

As the opensrf user, download the latest version of Evergreen trunk:

cd ~ && svn checkout svn://svn.open-ils.org/ILS/trunk Evergreen-trunk

As the root user, install the prerequisites:

cd /home/opensrf/Evergreen-trunk
make -f Open-ILS/src/extras/Makefile.install rhel

This will install a number of packages required by OpenSRF on your system, including some Perl modules from CPAN. You can say “no” to the initial CPAN configuration prompt to allow it to automatically configure itself to download and install Perl modules from CPAN. The CPAN installer will ask you a number of times whether it should install prerequisite modules - say “yes”.

As the root user, add /usr/local/lib/dbd to the system dynamic library path and make Linux recognize the newly installed libraries. Then restart PostgreSQL to avoid a problem where plperl.so cannot be found:
1. Create a file named /etc/ld.so.conf.d/eg.conf containing the following lines:
```
/usr/local/lib/dbd
```
2. Run the following commands:
```
ldconfig
chkconfig --levels 345 postgresql on #to start up postgres on boot
service postgresql initdb
vi /var/lib/pgsql/data/pg_hba.conf
```
3. Edit the pg_hba.conf file to contain only the following lines:
```
local all all trust 
host all all 127.0.0.1 255.255.255.255 trust
```
4. Restart the postgresql server:
```
/etc/init.d/postgresql restart
```

As the opensrf user, configure and compile Evergreen:

cd /home/opensrf/Evergreen-trunk
cat /usr/share/aclocal/libtool.m4 /usr/share/aclocal/ltoptions.m4 /usr/share/aclocal/ltversion.m4 /usr/share/aclocal/ltsugar.m4 /usr/share/aclocal/lt~obsolete.m4 >> aclocal.m4
./autogen.sh
./configure --prefix=/openils --sysconfdir=/openils/conf
make

As the root user, install the code. Set the STAFF_CLIENT_BUILD_ID variable to match the version of the staff client you will use to connect to the Evergreen server. Create a symbolic link named server in /openils/var/web/xul/ to the /server/ subdirectory of your staff client build.
```
cd /home/opensrf/Evergreen-trunk
make STAFF_CLIENT_BUILD_ID=current install
cd /openils/var/web/xul
ln -sf current/server server
```
Copy the example OpenSRF configuration files into place. This will replace the OpenSRF configuration files that you set up while installing and testing OpenSRF; you might want to backup the old files for troubleshooting purposes. Finally, change the ownership on the installed files to the opensrf user:
```
cp /openils/conf/opensrf.xml.example /openils/conf/opensrf.xml
cp /openils/conf/opensrf_core.xml.example /openils/conf/opensrf_core.xml
cp /openils/conf/oils_web.xml.example /openils/conf/oils_web.xml
chown -R opensrf:opensrf /openils/
```
As the postgres user on your PostgreSQL server, create the Evergreen database.

Issue the following commands on your PostgreSQL server:

createdb -E UNICODE evergreen
createlang plperl   evergreen
createlang plperlu  evergreen
createlang plpgsql  evergreen
psql -f /usr/share/pgsql/contrib/tablefunc.sql evergreen
psql -f /usr/share/pgsql/contrib/tsearch2.sql  evergreen
psql -f /usr/share/pgsql/contrib/pgxml.sql     evergreen
psql -f /usr/share/pgsql/contrib/isn.sql       evergreen

As the postgres user on the PostgreSQL server, create a PostgreSQL user named evergreen for the database cluster:
```
createuser -P -s evergreen
exit
```
Enter the password for the new PostgreSQL superuser ("evergreen")

As the root user, create the database schema and configure your system with the corresponding database authentication details for the database user that you just created; on most systems, <hostname> will be localhost and <port> will be 5432:

cd /home/opensrf/Evergreen-trunk
perl Open-ILS/src/support-scripts/eg_db_config.pl --update-config \
       --service all --create-schema --create-bootstrap --create-offline \
       --user <user> --password <password> --hostname <hostname> --port <port> \
       --database <dbname>

As the root user, set up Apache:

cd /home/opensrf/Evergreen-trunk
mkdir /etc/httpd/sites-available
cp Open-ILS/examples/apache/eg.conf       /etc/httpd/sites-available/
cp Open-ILS/examples/apache/eg_vhost.conf /etc/httpd/
cp Open-ILS/examples/apache/startup.pl    /etc/httpd/
chmod +x /etc/httpd/startup.pl
rm -f /etc/httpd/conf.d/welcome.conf
# Now setting up SSL
mkdir /etc/httpd/ssl
cd /etc/httpd/ssl
# Step 7
openssl req -new -x509 -days 365 -nodes -out server.crt -keyout server.key
# Step 8
vi /etc/httpd/sites-available/eg.conf

The openssl command cuts a new SSL key for your Apache server. For a production server, you should purchase a signed SSL certificate, but we can just use a self-signed certificate and accept the warnings in the staff client and browser during testing and development:
The last code line opens /etc/httpd/sites-available/eg.conf for editing:
Search for apache2 and replace with httpd
Comment out Allow from 10.0.0.0/8 and uncomment Allow from all (to enable access to the configuration CGI scripts from any workstation on any network - note that you must secure this for a production instance, preferably by locking down the allowed IP addresses and adding authentication, because you don't want just anyone adding and deleting libraries from your Evergreen instance!)
As the root user, edit /etc/httpd/conf/httpd.conf:
Search for apache2 and replace with httpd
Replace Include conf.d/*.conf with this line: Include /etc/httpd/sites-available/eg.conf
Change User apache to User opensrf
Change KeepAliveTimeout to 1
Change MaxKeepAliveRequests to 100

Add the following lines to the top of the LoadModules section:

LoadModule ssl_module modules/mod_ssl.so
LoadModule perl_module modules/mod_perl.so

For 64 bit installs only. Add the following lines the LoadModules section if they do not already exist:

LoadModule osrf_json_gateway_module /usr/lib64/httpd/modules/osrf_json_gateway.so 
LoadModule osrf_http_translator_module /usr/lib64/httpd/modules/osrf_http_translator.so

Update the prefork configuration section to suit your environment. The following settings apply to a busy system:

<IfModule prefork.c> 
StartServers       20
MinSpareServers    5
MaxSpareServers   20
ServerLimit      256 
MaxClients       256 
MaxRequestsPerChild  10000
</IfModule>

As the opensrf user install dojo:

cd /openils/var/web/js/dojo
wget http://download.dojotoolkit.org/release-1.3.2/dojo-release-1.3.2.tar.gz
tar zxf dojo-release-1.3.2.tar.gz
mv dojo-release-1.3.2/* .
rm -rf dojo-release-1.3.2*

As the root user, change ownership of the httpd logs directory:
```
chown opensrf /var/log/httpd
```
As the opensrf user, edit /openils/conf/opensrf_core.xml:
Edit /openils/conf/opensrf_core.xml to change the Jabber usernames and passwords as follows. I'm using XPath syntax on the left-hand side to indicate the position in the XML file:
- /config/opensrf/username = opensrf
- /config/opensrf/passwd = password for private.localhost opensrf user
- /config/gateway/username = opensrf
- /config/gateway/passwd = password for public.localhost opensrf user
- /config/routers/router/transport - first entry, where transport/server == public.localhost :
  - username = router
  - password = password for public.localhost router user
- /config/routers/router/transport - second entry, where transport/server == private.localhost :
  - username = router
  - password = password for private.localhost router user
We also need to specify the domains from which we'll accept and to which we'll make connections. If you are installing Evergreen on a single server and using the "private.localhost" / "public.localhost" domains, these will already be set to the correct values. Otherwise, search and replace to match your customized values.
Copy /openils/conf/srfsh.xml.example to .srfsh.xml in the home directory of each user you want to use to run the srfsh command line client for testing OpenSRF, and edit .srfsh.xml as follows:

domain is the router hostname (following our domain examples, private.localhost will give your srfsh access to all OpenSRF services, while public.localhost will only give you access to those OpenSRF services that are publicly exposed)
username and password must match your opensrf ejabber user for the chosen domain

logfile is the full path for a log file to which that user has write access

<?xml version="1.0"?>
<!-- This file follows the standard bootstrap config file layout found in opensrf_core.xml -->
<srfsh>
  <router_name>router</router_name>
  <domain>private.localhost</domain>
  <username>opensrf</username>
  <passwd>evergreen</passwd>
  <port>5222</port>
  <logfile>/tmp/srfsh.log</logfile>
  <loglevel>4</loglevel>
</srfsh>

As the opensrf user, make the cgi-bins executable, and add an environmental variable to opensrf's .bashrc file:

chmod 755 /openils/var/cgi-bin/*.cgi
echo "export PERL5LIB=/openils/lib/perl5:\$PERL5LIB" >> ~/.bashrc
. ~/.bashrc # inherit the new environment

Note: in a multi-server (brick) environment, put the ~/.bashrc modifications at the top of the file, before
```
[ -z "$PS1" ] && return
```
This will allow headless (scripted) logins to load the correct environment.

(Optional): Load translations such as Armenian (hy-AM), Canadian French (fr-CA), and others into the database to complete the translations available in the OPAC and staff client. See these instructions for details.

Starting Evergreen

As the root user, start the memcached and ejabberd services (if they aren't already running):
```
/etc/init.d/ejabberd start
/etc/init.d/memcached start
```
As the opensrf user, start Evergreen. The '-l' flag in the following command is only necessary if you want to force Evergreen to treat the hostname as 'localhost'; if you have configured opensrf.xml using the real hostname of your machine as returned by perl -ENet::Domain 'print Net::Domain::hostfqdn() . "\n";', you should not use the '-l' flag.
```
osrf_ctl.sh -l -a start_all
```
- If you receive the error message bash: osrf_ctl.sh: command not found, then your environment variable PATH does not include the /openils/bin directory; this should have been set by .bashrc when you logged in as the opensrf user, based on step 19 above, but you can manually set it using the following command:
```
export PATH=$PATH:/openils/bin
```
- If you receive the error message Can't locate OpenSRF/System.pm in @INC … BEGIN failed–compilation aborted, then your environment variable PERL5LIB does not include the /openils/lib/perl5 directory; this should have been set by .bashrc when you logged in as the opensrf user, based on step 19 above, but you can manually set it using the following command:
```
export PERL5LIB=$PERL5LIB:/openils/lib/perl5
```
As the opensrf user, generate the Web files needed by the staff client and catalogue and update the organization unit proximity (you need to do this the first time you start Evergreen, and after that each time you change the library hierarchy in config.cgi):
```
cd /openils/bin
./autogen.sh -c /openils/conf/opensrf_core.xml -u
```
As the root user, restart the Apache Web server:
```
/etc/init.d/httpd restart
```
If the Apache Web server was running when you started the OpenSRF services, you might not be able to successfully log in to the OPAC or staff client until the Apache Web server is restarted.

Testing connections to Evergreen

Once you have installed and started Evergreen, test your connection to Evergreen via srfsh:

Start srfsh and try logging onto the Evergreen server using the default administrator user ID and password:
```
/openils/bin/srfsh
srfsh% login admin open-ils
```
You should see a result like:
```
Received Data: "250bf1518c7527a03249858687714376"

------------------------------------
Request Completed Successfully
Request Time in seconds: 0.045286
------------------------------------

Received Data: {
   "ilsevent":0,
   "textcode":"SUCCESS",
   "desc":" ",
   "pid":21616,
   "stacktrace":"oils_auth.c:304",
   "payload":{
      "authtoken":"e5f9827cc0f93b503a1cc66bee6bdd1a",
      "authtime":420
   }

}

------------------------------------
Request Completed Successfully
Request Time in seconds: 1.336568
------------------------------------
```
If this does not work, it's time to do some troubleshooting.
- As the opensrf user, run the settings-tester.pl script to see if it finds any system configuration problems. The script is found at Open-ILS/src/support-scripts/settings-tester.pl in the Evergreen source tree. If the output of settings-tester.pl does not help you find the problem, please do not make any significant changes to your configuration.
- Follow the steps in the troubleshooting guide "checking for errors".
- If you have followed the entire set of installation steps listed here closely, you are probably extremely close to a working system. Gather your configuration files and log files and contact the Evergreen development mailing list for assistance before making any drastic changes to your system configuration.

Running the staff client on Linux

You can run the staff client on Linux using XULRunner.

Start up the staff client by passing the full path to the application.ini file for the source files of the local build of the Evergreen staff client. For example, if the source files for your Evergreen installation are in the /home/opensrf/Evergreen-trunk/ directory, you would issue the following command:
```
xulrunner /home/opensrf/Evergreen-trunk/Open-ILS/xul/staff_client/build/application.ini
```

Starting the Web server

Once you've started Evergreen and confirmed that a basic login attempt works, you can start up the Web server:

As the root user, test and start Apache. The restart command will ensure that it loads the new Evergreen modules even if Apache is already running:

apachectl configtest && /etc/init.d/httpd restart

If there are any problems with your configuration file(s), they will be displayed.

Stopping Evergreen

As the opensrf user, stop Evergreen:
```
osrf_ctl.sh -l -a stop_all
```

Setting up support for reports

Evergreen reports are extremely powerful, but some configuration is required.

Starting the reporter daemon

Once the open-ils.reporter process is running and enabled on the gateway, you have to start the reporter daemon. The reporter daemon periodically checks for requests for new reports or scheduled reports and gets them running.

To start the reporter daemon, run the following command as the opensrf user:

clark-kent.pl --daemon

You can also specify other options:

sleep=interval : number of seconds to sleep between checks for new reports to run; defaults to 10
lockfile=filename : where to place the lockfile for the process; defaults to /tmp/reporter-LOCK
concurrency=integer : number of reporter daemon processes to run; defaults to 1
boostrap=filename : OpenSRF bootstrap configuration file; defaults to /openils/conf/opensrf_core.xml

Stopping the reporter daemon

To stop the reporter daemon, you have to kill the process and remove the lockfile. Assuming you're running just a single process and that the lockfile is in the default location, perform the following commands as the opensrf user:

kill `ps wax | grep "Clark Kent" | grep -v grep | cut -b1-6`
rm /tmp/reporter-LOCK

Evergreen DokuWiki

Table of Contents