# Perlbug Installation instructions
# (C) 2000 Richard Foley RFI perlbug@rfi.net
#
# $Id: INSTALL,v 1.19 2001/08/16 21:03:53 mstevens Exp $
#
SHORT:
tar -xzf Perlbug.tar.gz
cd Perlbug
modify ./config/Configuration
perl Makefile.PL # PREFIX=Perlbug_installation
make
make test
make install
LONG:
Keep reading.
PREREQUISITES:
Perl :-)
MySQL installation from TCX
URI-1.09 or up from Gisle Aas
HTML-Parser-3.03 or up originaly from Gisle Aas
MailTools-1.13 or up from GBarr
Complete list of prerequisites in Makefile.PL
INSTALLATION:
Unpack the distribution in a suitable base level Perlbug directory.
Although it is recommended to create an appropriate system user,
'perlbug' for example, this is not necessary. However all file
and directory paths in these instructions will assume we are using
a dedicated 'perlbug' user account, and will be relative to his/her home.
In the slim chance that everything appears to be in a ~perlbug/perlbug
directory, stick it all in ~perlbug directly. ie; ~perlbug/Makefile.PL,
~perlbug/Perlbug, ~perlbug/Perlbug.pm, ~perlbug/sql, ~perlbug/web, etc.
The only files you should _need_ to edit are in the ~/config directory,
though, if using this application 'in anger', you will probably choose
to modify the ~/site files too.
Install suitable SQL database(currenly only Mysql supported, Oracle is
on the todo list), (make sure you can access the database with the
parameters you give to the ~/config/Configuration file).
Run the scripts in the appropriate ~/sql/* directories to create the
database, the default database user, all the tables, inserting some
minimal default entries etc. In particular, follow the instructions
in ~/sql/README
Set the Configuration file location in ~/Perlbug/Config.pm, this can
also be set 'on the fly' in the environment for testing, eg:
export Perlbug_Config=~/mytest/Configuration
Configure the ~/config/Configuration file which should be fairly
self-explanatory, but in particular ensure that:
database(username & password) are correct
directory entries all point to the correct locations
system::user is a real user on your system
directory::arch, spool(logs & temp) etc.
all exist and are writable by the appropriate user/s.
eg:
mkdir -p ~/perlbug/spool/archive
mkdir -p ~/perlbug/spool/logs
# etc.
chmod -R 777 ~/perlbug/spool
Setup mail forwarding to the appropriate scripts:
slurping the data: ~/scripts/bugtron (email/mailing lists)
query the data, admin interface: ~/scripts/bugmail (email)
Setup a cron job for regular tasks, clean up:
every week: ~/scripts/bugcron
Ensure ~/scripts/perlbug.cgi is executable by the webserver.
You'll also want to modify the ~/config/.htpasswd and ~/web/.htaccess entries
to enable the administration side of things via www, unless you're using
this in system::restricted=0 mode.
Create admins via the admins page at your config::web::domain
or adapt the ~/sql/generic/create_admin.sql script
For the default source and database data downloads to be operational
links or files will probably need to be installed from archive to
~/web/archive and ~/web/admin/archive
Voila! (hopefully :-)
COMMENTS:
The ~/spool(archive, logs, results, ranges, temp) directories are all
initially installed under the Perlbug root. These directories are liable
to fill up with volatile data and are prime candidates for moving to
another area (eg: /tmp/*), or disposing of periodically (eg: via cron).
Remember to reflect changes in the ~/config/Configuration file.
Any stray test bug data inserted in the database by 'make test' that
remains from an incomplete test run should be removed by ~/t/99_Test.t
Further info in the docs: 'perldoc Perlbug'
AUTHOR:
Richard Foley richard@perl.org perlbug@rfi.net 1999 2000