MINDSET CHANGES IN DATA MODELING

While technology is changing every day, the data sources and types also change. Nowadays, the data created every minute are massive and has different types (structured, semi-structured, and unstructured). For years, most organizations depended on relational databases (RDBMS), which only process structured data types. However, when the Big Data era started, it was required to find databases that could process the vast amount of data and different data types. Since big data has different kinds and data sources, NoSQL databases were the proper databases that could solve this issue. It can process and store that massive amount of data with the different data types.

Big companies such as Google and Amazon realized this, and they worked to invent new types of data models that can work with the other types of data. Also, they offered these models to be used for others. For example, Amazon created DynamoDB, a Key-Value and document database. According to the Amazon website, DynamoDB can process any volume of data within a single-digit millisecond. (Introduction to Amazon DynamoDB (1:01), 2021). In addition, Google built the Firebase database. According to the Google website, the Firebase database is a cloud-host database and deal with JSON files which are semi-structured data types (Firebase Realtime Database, n.d.). Being on the cloud makes the database more stable and available. Google said Firebase is a real-time database that allows the data for all clients connected to the database (Firebase Realtime Database, n.d.).

There are different types of NoSQL databases in the market that can be used by anyone who wants to use NoSQL rather than RDBMS. For example, Facebook created Cassandra, Bigtable was created by Google, Amazon, MongoDB, and Hbase created DynamoDB.

However, picking a NoSQL database need studying the organization needs to find the proper one. Each type of NoSQL database can process specific kinds of issues, and organizations must define their needs to choose the correct NoSQL databases that fit their needs.

In addition, Hadoop appears to work with NoSQL to solve these issues by making applying NoSQL easier and creating a system that makes using distributed computing much more straightforward. The following paragraph explains the difference between NoSQL and Hadoop.

NOSQL AND HADOOP

People sometimes mix between NoSQL and Hadoop. Hadoop is a software built by Apache Software Foundation and written in Java (GeeksforGeeks, 2019). Hadoop’s purpose is to process and store big data through distributed databases. Karanth, in the book “Mastering Hadoop: Go beyond the basics and master the next generation of Hadoop data processing platforms,” defines Hadoop as an open-source framework used in organizations for large-scale and distributed computing (2014). It is software that makes working with distributed computing processes harmonious (Lo, n.d.). On the other hand, NoSQL is a type of database that is not a relational database, and it can process vast amounts of data with a high ability to scale and deal with different types of data. In addition, NoSQL designed work with distributed computing databases and flexibility of schemas.

We can see how Hadoop and NoSQL have standard features from the previous introductions. Sinha and Shah, in an article, say that NoSQL and Hadoop have intersecting features, but they are not challengers. Both Hadoop and NoSQL are working with big data. Hadoop software allows using some types of NoSQL in the distributed computing system (Lo, n.d.).

To understand the relationship between NoSQL and Hadoop, we can say that NoSQL is the infrastructure that provides database architecture, and Hadoop is the software that runs and manage the NoSQL infrastructure. However, in some cases, organizations can run NoSQL databases without using Hadoop software; this depends on the data volume and velocity and other factors that determine the need for Hadoop (Langit, 2015).

However, because Hadoop can distribute the data through many nodes as needed to process and store the data, some organizations may not need that many advanced services and can use the devices they have using NoSQL with local software such as oracle (Langit, 2015).

CLOUD-BASED NOSQL DATABASE

Many different companies provide Cloud-Based NoSQL databases, for example, AWS from Amazon, Azure from Microsoft, and Google Cloud from Google. These services provide different NoSQL data models with customizable scales as needed. For instance, DynamoDB can be run using AWS, which allows the user to install Key-Value databases.

Furthermore, one of the best advantages of using Cloud-Based Databases is that they provide a service that allows the system to be scale as the data needed with very much easy steps compared to not cloud-based databases. In addition, Cloud-Based Databases make the creation of NoSQL databases much easier and faster, with no complicated settings that need to work on in the local NoSQL databases. All these services provide high stability and availability for the database, making the data processing work well.

Moreover, it can use the Hadoop system on the Cloud-Based services to install the configurations required in a few steps. All these features come with many cost plans. The could-Based database may be the proper approach for some organizations if they know exactly what they need and how to operate the cloud system. However, if the organization does not know the specific needs, a Cloud-Based Database could be not the right solution to use, and it can cost them a lot of money with no benefits for their business.