Home » BigData » The NoSQL Era – ‘Taking the leap’ – Part One

NoSql

Relational databases have been around for years. However over a number of years many of the biggest players in technology have found that the traditional relational databases are not good enough to handle the data problems of the modern internet. Therefore the likes of Facebook, google, amazon have created there own technologies which address the traditional problems, so a different term to data storage was coined ‘NoSql’. Although there is no specific definition for Nosql but there are common characteristics that define these technologies.

  • Non-relational
  • Scalability
  • Schema-less
  • Open-Source

Some of these characteristics address the common problems with traditional Relational databases.

Non-Relational & Schema-less

Although these are two different characteristics, they some what go hand in hand in a Relational database. Traditional relational databases have well defined schema’s that model the business, however these models are represented as relations in a tabular database. One of the biggest problems with this approach is that it becomes really inflexible when you want to change the schema specially when you have already set the relationships across the different tables.

Flexibility

If you know exactly what you want to model in your relational database and you know that your model will not change too much in the future then it is perfectly fine to go with a relational database. However this is not always the case as usually you want to have a little flexibility to change the schema, which is not easy in a relational database. This is where the non relational approach provides more flexibility, as it does not have to adhere to any explicit schema. Although there will always be an implicit schema as ultimately you will have to some business process that you need to model, which can be modelled at the application level. This provides the user with a lot of flexibility to extend and change the system as the business requirements changes.

Scalability 

One of the biggest problems with traditional databases is that they are difficult to distribute when there is a higher load of data and the user base increases. Due to the schema less none relational nature of nosql databases it is easier to distribute across multiple machines.

Polyglot persistence is the future

Relational databases will continue for a very long time due to its transactional capabilities. However with nosql technologies becoming more popular people should start to consider a mix of database technologies for your data storage.

Leave a Reply

Your email address will not be published. Required fields are marked *

Orgy