Use Orca to monitor system performance

We’ll be installing Orca on a Linux server. The server that we’ll be using runs Cent OS Linux version 5.3. Before you begin the installation of Orca you need to determine where you want Orca to store its installation binaries, raw data and the HTML reports. In this article we’ll be installing the Orca binaries and storing the raw performance data under the directory /opt/orca. The HTML reports will go under the Document Root of the web server, which is by default set to /var/www/html in most Red Hat Linux based computers. The default location for the installation binaries is set to /usr/local. So if you don’t specify anything else, Orca will install its files under that directory.
We would suggest using the latest snapshot of the development of Orca rather than its most recent stable release. You can grab this snapshot here. A recent snapshot release has also been bundled on the disc accompanying this magazine. Either way, once you have got hold of Orca, extract the archived file and ‘cd’ into the newly created directory using the following commands:

# tar -jxf orca-snapshot-r535.tar.bz2
# cd orca-snapshot-r535

There should be a number of files and directories in this folder. The instructions for the installation of Orca are contained in a file called ‘INSTALL’. We recommend that you read this before proceeding with the installation. There is also a directory called ‘packages’ which contains the installation files for all the Perl modules that Orca depends on. You can find the instructions on how to install these in the ‘INSTALL’ document. Make sure that you have all the Perl modules installed correctly before proceeding.
Let’s begin the installation. As mentioned earlier, we’ll be installing Orca in ‘/opt/orca’ and instructing it to place the HTML reports under ‘/var/www/html/orca’. Begin by creating these two directories, then proceed with the installation as shown below. Replace the directories in the commands shown below with the correct paths in your computer.

# mkdir /opt/orca
# mkdir /var/www/html/orca
# ./configure –prefix=/opt/orca –with-html-dir=/var/www/html/orca
# make
# make install

If all went well you should now have Orca installed on your computer. A few more steps and then we move on to the configuration bit. Under the main directory of the Orca installer there should be a file called ‘procallator_run_at_boot_using_chkconfig’. This will allow you to automatically start Procallator, which is the module of Orca that collects your data at boot if you are on a Red Hat based system, such as Fedora Linux or Cent OS. Use the following commands to get this up and running:

# make procallator_run_at_boot_using_chkconfig
# /etc/init.d/procallator start
Starting procallator:                                      [  OK  ]

Once you have successfully started Procallator, the data collection on your computer should begin. Procallator will begin collecting the system performance data based on the default settings. The Procallator configuration file, procallator.cfg, is in the ‘etc/’ directory under the Orca installation. In our case it is located at ‘/opt/orca/etc/procallator.cfg’. There are a number of configuration parameters that you can change here. You can change the frequency of data collection, the number items whose reading Procallator takes, the frequency with which Orca should generate performance graphs, as well as the admin email ID. We recommend that you at least change the admin email ID parameter from ‘warn_email root@localhost’ and replace it with your own email address. This way if something does wrong, you will be notified by mail.