Databases

The assigned reading from Access Science on database management systems gave a good overall description and background of databases and how they operate. As with most Access Science articles, it was a bit of a dry reading, and some of the terminology was far over my head. I thought the most interesting part was when the article discussed how databases are designed for one particular purpose. They gave the example that a database used to manage airline reservations could not hold data relating to payroll information. As far as the application to this class, this puts into context the complexity and extent of data storage and manipulation that would be needed to create, monitor, an maintain an intelligent building. Separate databases would be needed to monitor all aspects, and would have to be able to interact in order to have the building operate “intelligently,” or without the need for human influence.

Additionally, I read “Is the Relational Database Doomed?” by Tony Bain. He discusses the limitations of relational databases, citing the most prominent as the inability of the relational database to cope well with drastically increased data loads. In the technological society we live in, internet traffic in particular is difficult to predict and fluctuates rapidly. Their ability to scale is good on a single server node, but gets exponentially complex as that scaling spreads out over hundreds of thousands of nodes. The need for cloud services as led to a new breed of database management systems: key/value store. With key/value, no relationships are explicitly defined between different domains. Key/values also store all information pertaining to a particular item are stored within that item, and not referenced from a separately linked table. Because they are simpler, they scale better than the traditional relational database, but there are also more susceptible to bugs and errors.