Maintenance of Operational Systems
Challenge
"A system was developed for us, but it lacks accompanying documentation and our own IT staff is not familiar with maintenance procedures. How do we maintain or upgrade it in the future?"
DataArt's Method
In this situation, there are four major components: the system itself - in a production and test environment, the source code, documentation and staff expertise. The fewer components are present, the harder is the job for the new developer. Still, in many cases, it's possible to analyze and 'open' the system with only one component present – the system itself. DataArt combines practical experience with the fundamental knowledge of computer science principles in analyzing complex systems. The process adheres to the following rules:
"Exploration"
Based on the available information, we use advanced blackbox and whitebox analysis methods to probe the system. Input and output files are scrutinized and the environment is altered to 'scope' possible behavioral changes. Usually, the exploration is deployed using two types of developers:
- 'Hackers' – brilliant investigators who think out of the box and find original and creative ways to uncover missing details;
- 'Documentators' – systematic, focused individuals, who ensure that nothing is overlooked or interpreted incorrectly. They usually make excellent project managers when the project moves into production;
"Experimenting"
Once our research team has found ways to modify or integrate the system, a multi-threaded testing stage is conducted, ensuring that alterations are always in-line with expectations and that none of the previously working functions are broken.
"Documentation"
In parallel with the first two phases, this step provides an extensive information library, covering each aspect of the system. The library includes regular technical documentation, along with code samples and prototypes. All existing documentation is brought to the client's standard formats.
"Production"During this phase, DataArt or client teams have enough information to make reliable decisions and set realistic goals for the system.
DataArt's Resources
PeopleDataArt has successfully completed over 20 'exploration' projects and our experts are available for new challenges. It is the combination of advanced education and real-life projects that makes a researcher capable of solving tasks of this magnitude.
TechnologyOur Software Development Center maintains a large multi-platform R&D server farm that enables us to replicate virtually any production environment or system architecture running on Windows, UNIX and mobile platforms.
|
