Free Online Courses for Software Developers - MrBool
× Please, log in to give us a feedback. Click here to login
×

You must be logged to download. Click here to login

×

MrBool is totally free and you can help us to help the Developers Community around the world

Yes, I'd like to help the MrBool and the Developers Community before download

No, I'd like to download without make the donation

×

MrBool is totally free and you can help us to help the Developers Community around the world

Yes, I'd like to help the MrBool and the Developers Community before download

No, I'd like to download without make the donation

NoSQL vs NewSQL: BigData in Details

In this article we will see a complete discussion regarding these two new BigData methodologies: NoSQL and NewSQL, their main features and principles.

Today, information is vital for organizations that have multiple data sources and systems to store them. However, with the use of relational databases and normalization, which principle is data consistency, there is a big problem that slows the recording and retrieval because the number of data entered in these databases is becoming greater, turning the data analysis worse.

With the evolution of Web 2.0, the proliferation of social networks and the intense interaction with users, every day more and more data is inserted and retrieve. High-speed analysis is almost impossible since the networks do not give this support. NoSQL is designed especially for distributed architectures and concepts of non-normalization and not using a query language like SQL. But with its rampant use, some problems were encountered, such as lack of transactions use, lack of SQL queries and complex structure because of the lack of proper modeling. This led to a new proposal: the NewSQL. This is designed to extend the benefits of relational model for distributed architectures and improve the performance of SQL databases. There is no longer a need to increase server capacity by making use of horizontal scalability. In this article, I present the main characteristics of the NoSQL and NewSQL databases and their use in the present scenario.

NoSQL

The main features of NoSQL databases is the horizontal scalability based on a distributed memory type architecture with replication and data fragmentation on different servers, allowing support to a large number of I/O operations (Input/Output) per second. Most of these databases do not support the transactional ACID properties to achieve performance and greater scalability. The DBMS NoSQL operate on the relative performance and complexity of the model, always tending to increase the first. Let's see six of the most important features of NoSQL databases:

  • Increased performance of simple operations when the number of nodes is increased;
  • Replication and distribution of data on different nodes;
  • Instead of SQL, a simple communication protocol with the DBMS;
  • A concurrency control model more "relaxed" than those used in traditional RDBMS, called "late consistency";
  • Efficient distribution of the contents and use of RAM for data storage;
  • Dynamic adding attributes to records that already exist in the database (for non-obligation to have fixed schemes).

NoSQL databases hit the market as an alternative to meet the needs of applications that traditional DBMS found it difficult to meet because of the problems of scalability and availability. However, these solutions are already facing some problems, because they lack a unified query language such as SQL, and require a manual setting of the consultations, which may not be a feasible task for those who do not know how to encode these solutions. In addition, over the years we had a lot of investments in algorithms and ways to improve SQL compilers to generate execution plans of optimized queries. Thus, to develop new methods of access, we lose this experience. And as the NoSQL supports only delayed data consistency for applications where data accuracy is critical, not having consistency guarantees can be a big problem. NoSQL databases can be classified into different types, according to the proper use of them for a given task. They can be classified as:

  • Key-value pairs: System that stores indexed values for later retrieval through keys;
  • Oriented Column: Unlike DBMS that stores information in highly structured tables, specifically in rows and columns, this type of system has an extendable table minimally related data. Each row of this table can have its own set of columns, which need not be equal to the others. Scalability is done through the distribution of rows and columns by nodes, initially fragmenting the columns of a table and then fragmenting the resulting lines of these tables by the primary key, with fragmentation by interval;
  • Graph databases: They are directly related to graphs data models, with the idea to represent the data as structures that generalize the notion of graphs. This model is proposed to when the topology of the data is more important;
  • Document-oriented: These databases use the concept of self-contained and self-describing data and documents. This implies that the document itself already defines how it should be presented and its structure defines how they should be stored.

NewSQL

The NewSQL databases seek to provide the same improved performance and scalability of NoSQL systems, not giving up the benefits of traditional databases, the SQL and ACID properties. Mike Stonebreaker, founder of VoltDB (one of this new model databases), highlighted the advantage of NewSQL databases for providing consultations in real time as well as increasing processing power. According to Mike, there is a great cost when not using SQL, and excessive work required developers to compensate for their absence. Unlike traditional DBMS, which were considered solutions for any type of application, NewSQL use a different strategy, where each new developed system aims to meet a specific market need and seeks to achieve it separately, ending with the old concept of having a single system that fits any application, so that the databases are experts for a purpose, not generating more absurd number of unnecessary functions and behaviors for a particular application. Let's see some features of a DBMS NewSQL:

  • SQL language as a way of interaction between the DBMS and the application;
  • Support for ACID transactions;
  • Non-blocking concurrency control, so that reads and writes do not cause conflict with each other;
  • Architecture that provides higher performance per processing node;
  • Scalable architecture with distributed memory and ability to function in a cluster with a large number of nodes.

Table 1shows a comparison of the characteristics of relational databases (RDBMS), NoSQL, and NewSQL with their abilities and strengths.

Characteristic

RDBMS

NoSQL

NewSQL

ACID

Yes

No

Yes

OLAP/OLTP

Yes

No

Yes

Data analysis

Yes

No

Yes

Scheme rigidity

Yes

No

Maybe

Flexibility data format

No

Yes

Maybe

Distributed computing

Yes

Yes

Yes

Up scale

Yes

Yes

Yes

Performance data with growth

Fast

Fast

Very Fast

Performance overhead

Large

Moderate

Mínimo

Popularity/Community support

Large

Growing

Growing slowly

Tabela 1. Comparing RDBMS, NoSQL and NewSQL

From this table we can see that NewSQL is a improvement ofRDBMS, using the best of both worlds. It totally kept the features of the standard databases, implementing the innovations ofNoSQL. In items in which it is "Maybe" is because it depends on which NewSQL database we are using, because, as we have seen, each database has a specific purpose. There are some NoSQL experts databases with the power of NewSQL, such as MariaDB (Developed by the creator of MySQL).

Big Data and its relationship to NoSQL and NewSQL databases

Big Data applications originated with the arrival of Web 2.0, developed with great speed, and its proposal is to give useful information quickly to the user or allow the company to make good decisions at the business level.

With the use of these applications by Internet giants like Google, Facebook and Youtube, the concept of BigData became even more popular , being considered by the Gartner team a technology that will have great importance in the modernization of information management with strong impacts on infrastructure information and corporate governance. According to Gartner, Big Data supports processing solutions for a variety of new and existing data, bringing real benefits to the business. Still, the processing volume and variety of data is still a technical solution, unless it is part of business goals.

The essential aspect of Big Data is not the technical skill of data processing, but the benefits that can be achieved by an organization using analysis of Big Data.

The problem today is that the data generated in various ways, generating a large body of data for analysis, giving rise to the problem with relational databases by no ability to process so much information with high availability, optimization and scalability. Thus, the need for more efficient database systems for these tasks is high, encouraging the use of NoSQL databases and NewSQL.

Conclusion

Both NoSQL and NewSQL are now part of all the softwares that we have available at the market and we must adapt to them if we want to stay into the market. It’s extremely important to understand their features and experience them to make the best use in general.



Web developer and passioned for web design, SEO and front end technologies.

What did you think of this post?
Services
[Close]
To have full access to this post (or download the associated files) you must have MrBool Credits.

  See the prices for this post in Mr.Bool Credits System below:

Individually – in this case the price for this post is US$ 0,00 (Buy it now)
in this case you will buy only this video by paying the full price with no discount.

Package of 10 credits - in this case the price for this post is US$ 0,00
This subscription is ideal if you want to download few videos. In this plan you will receive a discount of 50% in each video. Subscribe for this package!

Package of 50 credits – in this case the price for this post is US$ 0,00
This subscription is ideal if you want to download several videos. In this plan you will receive a discount of 83% in each video. Subscribe for this package!


> More info about MrBool Credits
[Close]
You must be logged to download.

Click here to login