NAME

Karma - Installation Guide


Standard Install

Karma finally has a standard perl install. It comes with a Makefile.PL which you use to build a makefile for your system. You can do that as follows:

$ perl Makefile.PL

The resulting ``Makefile'' can be used to make, test, and install karma.

Once you've installed karma on your system, you may want to create a karma user in your database. There are scripts in the ``sql'' directory for creating the user and prompting for a password. In addition if you will be using alertlog and os monitoring, you'll want to create the karma objects in the karma user schema. There is a script in the ``sql'' directory to create these objects as well.

Once you have everything installed, the

Quickstart Guide

should explain to you how to get started. See the

Configuration Guide

for more information on how to create a

read-only user

for karma in your database, to

edit the karma.conf file

, or to use the karmagentd to

monitor the alertlog

and uptime statistics on your remote database.


Using Karma with a webserver

Yes, you can use karma with a webserver, but there isn't too much you need to do on the karma end of things. Basically karma generates a set of html pages. These by default go into ``doc_root'' in the karma software directory, but this path can be changed via the -k option to karmad when you start it or the ``doc_root'' directive in your config file. Specify a valid path, and it will write the files there. Be sure to move the existing karma documents from doc_root to that location, and create directories named ``docs'',``help'', ``images'', and ``info'' there as well. In short, you should copy all the files from the install doc_root except maybe karma.html, which will be recreated.


Installing Perl

Installing perl has become *MUCH* easier over the years, so don't be so afraid of installing it on your own system. You can install it from

source

binaries

Installing on linux systems (most of which come with perl anyway) is often as simple as:

$ gunzip < stable.tar.gz | tar xvf -

$ cd perl5.005_03$ ./configure

$ makeand as root$ make install


Installing DBI

It's pretty straightforward on most machines where Perl is already installed. Download the latest DBD-Oracle from your favorite

CPAN Site

, then follow these steps:

$ gunzip < DBI-1.13.tar.gz | tar xvf -

$ cd DBI-1.13$ perl Makefile.PL

$ make$ make testthen as root:

$ make install


Installing DBD-Oracle

It's pretty straightforward on most machines where Perl is already installed. Download the latest DBD-Oracle from your favorite

CPAN Site

, set your environment variables ORACLE_HOME, ORACLE_SID, be sure $ORACLE_HOME/bin is in your path, and that $ORACLE_HOME/lib is in your library path then followthese steps:

$ gunzip < DBD-Oracle-1.03.tar.gz | tar xvf -

$ cd DBD-Oracle-1.03

$ perl Makefile.PL

$ make

$ make test

then as root:

$ make install


Installing Oracle

Of course, none of this would be complete without an Oracle database to connect to. Installing Oracle, however, is really beyond the scope of this document. You can find useful documentation at

http://technet.oracle.com

though you'll need to register first. It's free. In addition, you can find linux specific installation instructions here:

http://jordan.fortwayne.com/oracle/