{T³} 2. Trading System Architecture

And now, going concrete on the architecture domain:

Any software you build for the aforementioned purposes will have to cope with a large set of data, be highly available and quite connected. These I solved by the following:

Physical boxes:

Analytic database and application servers run on-premise.

Web server for reporting and general UI interaction resides in the Amazon cloud.

The mobile client is running thick on my tablet computer.


You have a variety of solutions to select from. Free solutions can be competitive. In my case, I bought MS SQL Server 2014 for the sake of development. It is not only a transact-SQL system, but a lot more – the relevant value-added components being SSAS and SSIS (called as such, before renamed).

Application server:

Again, you have a number of alternatives. Free solutions have their limitations but – so far – nothing as bad as could be called a show-stopper. So I selected Glassfish 4 over J2EE. This platform is free, easy to manage and reliable for my purposes. In order to get all the magic working, I needed to add the following libraries:

Mobile Client:

This component, as hinted before would be rather called a streamlining add-on, or productivity feature. It only makes sure that the end user of the system gets anything for human confirmation by the hand as efficiently as possible.


Finally, let’s mention the Integrated Development Environment I use. That is Eclipse, coming with the foundation offering plus all the proper additives for enterprise server handling (deployment, etc.) and Android development, added by myself.

Theoretical background: