Think Outside The Row And Column

These days you can’t talk about online business without hearing about big data and analytics. There is a lot to be gained from the data generated by social networks, shopping, streaming, gaming and other activity online. Where is all this data stored though?

Are you interested in how data is stored and managed? Read more about data governance.

Relational Databases Store Your Data 

The data everyone generates online is generally stored in databases. Relational databases are the most commonly used type and they contain linked tables. Each of these linked tables contain columns and rows.

The columns define different properties of an entry and rows indicate a single data entry. Columns and rows intersect at cells, with each cell containing a piece of data as part of the data entry row. The special column that links two or more tables is called the primary key and its cells must all contain unique data.

From these descriptions, it’s clear that a majority of the data we generate and access on the internet can be easily stored in such databases.

The Pros and Cons of Relational Databases

Relational databases and the systems used to manage them are very stable and have a number of properties that make them very popular.

One property is the inbuilt ACID compliance that is great for ensuring financial transactions are completed correctly and securely. This is supported by the very well established language that is used to build and interact with these databases, Structured Query Language (SQL). Though this is technically a standard and so there are many versions associated with their own Relational Database Management System (RDBMS).

However, this highly structured nature of relational databases is also their disadvantage as they are rigid to the defined nature of the columns in their tables. So, if you have a new set of data that doesn’t fit your tables’ parameters, it will be difficult to include. Also, such databases are difficult to scale, needing ample investment in infrastructure every time the databases increase in size.

The Alternatives to Rigid SQL Database Systems

The disadvantages of relational databases, especially the difficulty in scaling has created demand for NoSQL databases. NoSQL stands for ‘Not Only’ SQL and this includes different database structures other than the table format.

The four currently popular NoSQL databases include document databases such as MongoDB, graph databases such as Neo4j, key-value stores such as Redis and Column databases such as Apache Cassandra.

Graph Databases (Neo4j)

The main idea behind graph databases is that the relationships between data points are as important as the data. So graph databases have two major parts, the data which is stored as nodes and the relations between the data which are stored as edges. When you look at a visual representation of such data bases, they look like large flexible ball and string diagram.

Neo4j is the most popular implementation of this property graph model and uses the Cypher language to interact with its databases. This method of data storage and accessing that is not strictly structured makes storage and accessibility highly scalable without needing powerful/ expensive hardware.

The simple design of the data model makes graph databases intuitive yet very flexible to manage. Large data sets can have new properties (edges) added at almost any time and with minimal hardware and time required.

Document Databases (Mongodb)

Document databases bring flexibility of data by storing it discretely in groups of key-value pairs called documents. The key-value pairs tend to have a similar format to JSON and this makes the schema of each document very flexible. Documents with a similar purpose are stored in a collection. This collection is the equivalent of the relational database table.

Since the data is store in this independent fashion, it is much easier to quickly store data from many sources without having to worry about a formal schema as long as it is in an appropriate collection.  Querying large amounts of data is also faster than in relational databases as it is stored in discrete collections that are not weighed down by complex join functions.

Are You Making the Most of your Data?

Today’s data, weather relational or not, is only growing.  Make sure your company is set up to handle the load and make the most of your data.

Are you looking to migrate your data from relational to NoSQL databases? Let the experts at ASB Resources walk with you every step of the way. Schedule a call with one of our experts today!