Cluster-Handbook/Software

Software
The topic of this section is installation and maintenance of software that is not present in the default repositories of Ubuntu. Different methods to install, update and integrate software into the environment will be presented.

The Problem
Installing Software on GNU/Linux systems usually consists of three steps. The first step is downloading an archive which contains the source code of the application. After unpacking that archive, the code has to be compiled. Tools like Automake assist the user in scanning the environment and making sure that all dependencies of the software are satisfied. In theory the software should be installed and ready to use after the three simple steps:



Almost every provided README file in such an archive suggests doing this. Most of the time however, this procedure fails and the user has to manually solve issues. If all problems are solved, the code should be compiled and installed  (the second and third step). There are alternative projects to Automake like CMake and WAF that try to make the process less of a hassle. The process of installing and integrating software into an existing environment like this can take quite some time. If at some point the software has to receive an update, it is not guaranteed to take less time than the initial installation.

The root of this problem is that different distributions of GNU/Linux are generally quite diverse in the set of software that they provide after installation. This means if there are 5 different distributions, and you want to make sure that your software runs on each of them, you have to make sure that your software is compatible to potentially:


 * 1) 5 different versions of every library your software depends on.
 * 2) 5 different init systems (which take care of running daemons).
 * 3) 5 different conventions as to where software has to be installed.

Tools like Automake, CMake and WAF address this problem, but at the same time the huge diversity is also the reason why they are no 100% solutions and often fail.

Distribution Packages
Instead of trying to provide one archive that is supposed to run on every distribution of GNU/Linux it has become common to repackage software for each distribution. The repositories of Debian contain thousands of packages, that have been packaged solely for one release of Debian. This makes the installation and updating a breeze, but causes severe effort for the people who create these packages. Since Ubuntu is based on Debian most of these packages are also available for Ubuntu. These packages are pre-compiled, automatically install into the right location(s) and provide init scripts for the used init system. Installation of them is done by a package manager like apt or aptitude, which also manages future updates. There are only two minor problems:


 * 1) Software is packaged for a specific release of the distribution. For example Ubuntu 12.04 or Debian Squeeze. Once installed, usually only security updates are provided.
 * 2) Of course not every software is packaged and available in the default repositories.

This means, if you are using the latest long term support release of Ubuntu, most of the software you use is already over one year old and has since then, only received security updates. The reason for this is stability. Updates don't always make everything better, sometimes they break stuff. If software A depends on software B it might not be compatible to a future release of B. But sometimes you really need a software update (for example to get support for newer hardware), or you just want to install software that is not available in the default repositories.

Personal Package Archives
For that reason Ubuntu provides a service called Personal Package Archives (PPA). It allows developers (or packagers) to create packages aimed at a specific release of Ubuntu for a specific architecture. These packages usually rely on the software available in the default repositories for that release, but could also rely on newer software available in other PPAs (uncommon). For users that means, they receive software that is easy to install, should not have dependency problems and is updated frequently with no additional effort. Obviously this is the preferred way to install software when compared to the traditional self compiling and installing.

How are PPAs used?
PPAs can be added by hand, but it is easier using the command. That command is provided by the package.

The command  is used with the PPA name preceded by. An important thing to know is, that by installing such software you trust the packager that created the packages. It is advised to make sure that the packages won’t harm your system. The packager signsthe packages with his private key and also provides a public key. That public key is used by  to make sure that the packages have not been modified/manipulated since the packager created them. This adds security but as it was said before it does not protect you from malicious software that the packager might have included in the software.
 * Listing 2.1: Install python-software-properties for the command add-apt-repository.

After adding the repository you have to call  to update the package database. If you skip this step the software of the PPA won't be available for installation.
 * Listing 2.2: Add the ppa repository.

Now the software can be installed using. It will also receive updated when  is called. There is no extra step required to update software from PPAs.
 * Listing 2.3: Update the package database.


 * Listing 2.4: Install the desired packages.