Why is Apache Spark Becoming More Popular?

2015 may be known as the year Apache Spark really came into its own. Spark was first launched back in 2009, and its popularity has been steadily growing ever since. However, in just the past year, Spark has exploded onto the scene. Perhaps this shouldnt be a surprise since the signs were there beforehand. After all, at the tail end of 2014, Apache Spark finally surpassed Hadoop in interest based on statistics from Google Trends. This doesnt necessarily correlate directly with Sparks impact on the big data community, but the trend is clear, and its become a full blown phenomenon in 2015.

While theres no denying the upward trend of Sparks popularity, some may be wondering why more people and companies are paying so much attention to the technology. Understanding the reasons behind Sparks rise can help in predicting the trajectory of big data solutions as well as companies that make use of the technology.

Many large organizations have seen the value that Spark can bring to their businesses. IBM, for example, recently made a $300 million dollar investment in Apache Spark. Many businesses that are focused on big data have made the switch from Hadoop to Spark. The number of contributors to Spark has also reached impressive levels. Apache Spark has around 600 contributors at the moment, making it the most active Apache project around, which is impressive considering how many projects are part of the Apache Software Foundation.

So why have so many companies decided to make Spark part of their big data operations? Theres no single explanation, but put simply, Spark can make big data easier to work with and handle. Many companies prefer to use the platform for crunching massive amounts of structured and unstructured data.

For example, Spark supports in-memory parallel processing power which can run programs much more quickly than alternative methods. At the same time, data sources can be blended and harmonized all at the same time, saving a lot of resources and time in the long run. By processing big data faster and more easily, companies can finally be more successful in their use of big data technologies and avoid many of the big data pitfalls businesses become ensnared in. Is it any wonder that according to a recent Databricks survey, the number of attendees to this years Spark Summit increased by 156 percent, and the number of companies represented grew by 152 percent?

Sparks advantages go even further than that. The technology is open source, which tends to outperform traditional software. Open source options have been improving greatly in the past few years, providing for more innovation and higher quality. That makes it an excellent choice for companies to embed in their software when the need to analyze big data arises. Equally valuable to Sparks popularity is the tools that the technology includes, like machine learning, interactive SQL, and real-time processing. These capabilities are essential for properly analyzing real-time data as it is collected and transmitted from connected devices. This makes Spark an integral part of the Internet of Things (IoT) as it takes off.

Apache Spark is highly versatile as well. While many view Spark as the technology that will dethrone Hadoop, it doesnt have to be that way. In fact, Spark is fully compatible with Hadoop while still being able to run independently from it. The Hadoop Spark combination contributes to the processing of big data in many helpful ways. This makes a full switch from using Hadoop to Spark unnecessary, but in cases where a company only wants to use Spark, the change can be a smooth one without the usual hiccups and problems a typical transition entails.

Apache Spark has caught on with businesses and developers in a short amount of time. Considering the added emphasis on big data, the cloud, and the Internet of Things in recent years, this type of popularity surge should be expected. Spark has a lot to offer to companies willing to use it. It features major improvements, user-friendly features, and a vast array of tools at a developers disposal. The more organizations want to use big data, the more it becomes clear that Spark will be part of that future.