Server Building Blocks

Open Source packages provide an excellent foundation for building a server. The most important ones for the LongJump architecture are listed below.

• The Linux operating system.

• MySQL and PostgreSQL databases.

• The Flightdeck-UI MVM monitoring tool

• The GNU C++ compiler and libraries.

• The Python interpreter and libraries.

Overall, the solution based on the above components enhances reliability, provides a secure platform, and drives down costs. Database requirements are readily met by the lightweight MySQL (best for smaller amounts of data, and optimized for read access) or the powerful PostgreSQL (a highly sophisticated product that is a genuine alternative to Oracle).

The Flightdeck-UI MVM product is very useful for server monitoring. MVM includes a graphical editor with theme support, which allows it to be adjusted for a particular application. Figure 2 shows two small examples of MVM control panels (the examples are monitoring outdoor temperatures in several major cities around the world).

The primary programming languages on the server side are C++ and Python. C++ is one of the world's most popular programming languages, and features very high performance with a rich Object Oriented feature set. Building critical server components with C++ reduces costs by requiring fewer servers to handle a given number of users. This is due to increased performance of each server. A well-designed C++ program further lowers costs and improves system stability by making efficient use of the hardware resources.

Components where performance is not critical should be built with a safe, easy-to-use language. This is especially helpful in those subsystems that will change frequently in order to continue customizing the application. Such an approach makes it easy for developers to maintain and enhance the system. The Python programming language is an excellent choice for complex components that are not performance sensitive.

One of the most user-friendly programming languages in use today, Python nevertheless supports a powerful set of Object Oriented features that make it highly suitable for complex projects. Developers with varying backgrounds (including C++, C, Perl and Java) will find Python a very easy language to work with. Python is widely used in many organizations such as NASA, Lawrence Livermore National Laboratories, Red Hat, Industrial Light & Magic, Yahoo and Google.