By Kato Mivule
Database Systems
Before we take a look at what improvements could be implemented to enhance the current database architectures of such Online Social Network (OSN) applications as Facebook and Twitter, lets take a look at the distinctiveness of these Database architectures.
There are different types of Database architectures when it comes to Database Systems. The most notable architectures include the 1-tier, 2-tier, 3-tier, and n-tier architectures. In this type of architecture, there is a Database, Web Server, and Browser or Presentation Application. [1]
Database System architectures include the Presentation Layer which is responsible for user interaction and input validation. The Presentation Layer is responsible for presenting data to the user on the browser or presentation application.[2]
Secondly, we have the Application Processing Layer that is also referred to as the Business Logic Layer, responsible for serving as the middleman between the Application Layer and the Database Layer. Business functions in the Application Layer include search functions, Content Management, and Browse functions.[3] [4]
Thirdly, we have the Database Layer. It is in this layer that Data is stored, including the data structures and constraints. The Database layer carries out duties ranging from storage of data, security functions, and computer backup of data. [5]
In all, Database Architectures are an implementation of the Client-Server Architecture in which the Application is split into user interaction processes that run on the Client Side on a user’s PC for instance and the main process or programs runs on the Server side. The Client side is mainly used by the user and is the PC while the Server side runs the Applications and can be accessed my many Clients at the same time being connected together over a Communication Network. [6]
It is in the context of the Client-Server architecture that various Database Architectures are built, from the 1-tier, 2-tier, 3-tier, and n-tier architectures. In the 1-tier Database Architecture, Presentation, the Application Layer, and Database Layer all reside on one computing device, that being the PC. In this case the user has full access to both the Presentation, Application, and Database. The Database system is not shared with others. [7]
In the 2-tier Database Architecture, the Database System is divided into two parts running on two separate machines, with the Presentation Layer running on the Client side, while the Application and Database Layers run on the Client side. In this type of architecture, many Clients can access the Server which in this case is called the database Server. [8]
It is in the 3-tier Database Architecture that the Database system is split into three parts running on three machines, the Presentation Layer running on the Client side, the Application Layer running on an Application Server or Web Server in the case of OSN Applications like Facebook, and the Database Application running on a separate Database Server. The users, in this case the Clients will request access and services to the Database Server through an intermediary called the Application Server. [9]
This is the under-laying architecture of most OSN applications like Facebook and Twitter. There is great room for improvements, future enhancements, and optimizations. There is a great increase in the use of Mobile Computing Devices and this means that the Architectures of most Databases will have to cater for the exponential utilization of Mobile Devices such as PDAs, Kindles, iPhones, iPads, etc. The improvements will not as much focus only on the Database Layer but on the Presentation and Application Layer.
First, with the with Database Layer, improvement in resources will be critical. Right now Facebook boasts of about 500 million users, [10] with such a scenario, if all these users get Mobile Computing Devices, then hardware and software resources at the Facebook Database Layer would need to be greatly improved in order to accommodate 500 million requests in a given day, or in the worst case scenario, 500 million requests to the database server per second. However, it is at this Database Layer, that improvements could be implemented to not have a single database but multiple databases running various user accounts in a distributed model, as such sharing resources in a worst case scenario when 500 million requests are made per second.
At the Presentation and Application Layers, improvements would need to be made to present data in fitting with the display parameters of most mobile computing devices. Software could be made intelligent to sense the type of computing device that the client is using. Such Intelligent Sensors would be implemented at both the Application and Presentation layers. When the client connects to the Application Layer, the client sensor sends information to the Application sensor, and then data is presented in specification to the client displays.
Security is another critical feature that would regularly need improvement. With the advent of OSNs, a lot of Personal Identifying Information(PII) is stored on the web. This means that malicious attackers present a great deal of risk in stealing PII such as Social Security Numbers, Credit Card Numbers, Date of Birth information, etc.
To deal with this problem, since user accounts on most OSNs are personalized, we could deploy Intelligent Security Agents at the Application and Presentation Layers that would work in concert with a Client Mobile Computing Device should they detect any client PII online, communication could then be sent to the Client warning of a PII data leak. This approach could even be used for large organizations in presentation of Statistical Data.
The Intelligent Security Agent would first scan for numerous OSNs at the Presentation and Application layers for any PII related to the client for instance or any linking of such PII data to the soon to be released data. If the Intelligent Security Agents return positives, then the large organization would have to employ various algorithms to further scramble their statistical data to the extent that malicious attackers would not be able to infer or reconstruct any client Identities.
In conclusion, improvements in data presentation, dynamic database resource allocation, and security are critical in any database architecture today.
References
[1] “Distributed Application Architecture”. Sun Microsystem. http://java.sun.com/developer/Books/jdbc/ch07.pdf. Accessed: 09-Sep-2010.
[2] “Three-Tier Architecture | Linux Journal.” [Online]. Available: http://www.linuxjournal.com/article/3508. [Accessed: 09-Sep-2010].
[3] Wenyin Liu, Qing Li (Ph.D.), Qing Li, Rynson W. H. Lau. “Advances in web based learning: ICWL 2006 : 5th international conference, Penang, Malaysia, July 19-21, 2006 : revised papers, Volume 4181 of Lecture notes in computer science, LNCS sublibrary: Information systems and applications, incl. Internet/Web, and HCI. Springer, 2006, ISBN 3540490272, 9783540490272, Page 90
[4] “Multitier architecture – Wikipedia, the free encyclopedia.” [Online]. Available: http://en.wikipedia.org/wiki/Multitier_architecture. [Accessed: 09-Sep-2010].
[5] Ashutosh Deshmukh. “Digital accounting: the effects of the Internet and ERP on accounting”. Idea Group Inc (IGI), 2006. ISBN 1591407389, 9781591407386, Page 21
[6] David G. Messerschmitt. “Understanding networked applications: a first course, The Morgan Kaufmann series in networking”. Morgan Kaufmann, 2000. ISBN 1558605371, 9781558605374, Page 140.
[7] Ricardo Jardim-Gonçalves, Rajkumar Roy, Adolfo Steiger-Garção. “Advances in concurrent engineering: proceedings of the 9th ISPE International Conference on Concurrent Engineering, Cranfield, United Kingdom, 27-31 July 2002″, Taylor & Francis, 2002. ISBN 9058095029, 9789058095022, Page 974.
[8] Lee Chao. “Database development and management, Foundations of database design series”. CRC Press, 2006. ISBN 0849333180, 9780849333187, Page 341.
[9] Mike Barley, Michael Wayne Barley, Nik Kasabov, Nikola K. Kasabov. “Intelligent agents and multi-agent systems: 7th Pacific Rim International Workshop on Multi-Agents, PRIMA 2004, Auckland, New Zealand, August 8-13, 2004 : revised selected papers, Volume 3371 of Lecture notes in computer science: Lecture notes in artificial intelligence”. Springer, 2005. ISBN 3540253408, 9783540253402, Page 318.
[10] “Facebook reaches 500 million users | Technology | The Guardian.” [Online]. Available: http://www.guardian.co.uk/technology/2010/jul/21/facebook-500-million-users. [Accessed: 09-Sep-2010].
