Databaseless design

Databaseless Design

Databaseless Design, or Database-less Design (DLD) is the design of web sites that provide dynamic data but do not store and retrieve that data from a database. These designs can be very sophisticated, providing functionality traditionally provided by database driven designs, such as user accounts and profiles, content and content management systems, as well as statistical tracking and reporting tools.

Why lose the database

There are a number of advantages to using database, relational or otherwise, in the design of a solution, but there are also a number of disadvantages.^[1] Disadvantages include the resulting complexity of relational schemas and the queries that are used to access the actual data, the technology stack required to run a database server and the expertise to keep it maintained, the cost of licensing and the added complexity of keeping the database secure. These responsibilities are normally spread across a number of job roles and members of staff from database administrators to the developers who design the software solutions, whereas with a databaseless design, often a single member of staff can implement the whole solution, bringing the costs immediately down.

Typical Technologies used

Typically, the storage and retrieval layer of an n-tier solution utilises replacement technologies for the database. These technologies often include XML^[2] as a datasource and sometimes even CSV files. XML is verbose, whereas CSV is an extremely efficient file format for lots of data.

In ASP.Net, there is the XMLDataSource^[3] which provides a quick and easy web control for serialising and deserialising data from XML files such that it can be used as a datasource for other components, such as datagrids, dropdown lists, and so on.

Challenges with databaseless designs

Concurrency is a challenge to be solved with any databaseless design, as is redundancy and backups. Performance, too, could become an issue if some sort of caching of the datafiles is not utilised, and for data that needs to be secure then some sort of encryption must be put in place. These challenges have to be carefully considered in the design in order to make any solution that does not utilise a database robust.

References

1. ^ http://www.cl500.net/pros_cons.html
2. ^ http://en.wikipedia.org/wiki/Xml
3. ^ http://msdn.microsoft.com/en-us/library/system.web.ui.webcontrols.xmldatasource.aspx