A new report from Dresner Advisory Services suggests that managed Big Data analytics platforms based on instances of Hadoop delivered via the cloud gained significant traction in 2017.
The research firm’s 2017 Big Data Analytics Market Study finds that instances of Hadoop curated by Cloudera, Hortonworks, and MapR Technologies still lead in terms of overall adoption. But now close behind are instances of Hadoop delivered as a managed service, namely Amazon EMR, Microsoft Azure HDInsight, IBM BigInsights, and Google Dataproc.
Overall, the survey finds 53 percent of respondents have implemented a Big Data analytics platform. That’s up from 17 percent in 2016. Another 40 percent says they expect to implement a Big Data analytics platform in the next two years.
Big Data analytics use cases
The most popular use case for Big Data analytics platforms is to modernize an existing data warehouse, also known as building a data lake. Most data warehouses today reside in an on-premises environment, which is where vendors such as Cloudera, Hortonworks, and MapR have enjoyed most of their success. But, the rise of cloud-based platforms suggests data gravity is starting to shift. As more applications reside in a public cloud, it starts to make sense to use Big Data analytics platforms that are co-resident with those applications to analyze data. The cost of moving and securing large amounts of data is often prohibitive.
New study finds 53% of respondents have implemented a #BigData analytics platform, up from 17% in 2016
The Big Data trend managed services providers will want to keep an eye on in the coming year is the rise of real-time analytics. As more organizations invest in digital business transformation and Internet of Things projects, there is an increased need to act on the insights garnered by an analytics application in real time. Because of that requirement, interest in next-generation relational databases is on the rise. By comparison, most instances of Hadoop are targeted at more batch-oriented applications. At the same time, however, performance of applications running on top of Hadoop is expected to increase considerably in 2018.
Opportunities for MSPs
All told, MSPs should expect to see the amount of data being stored in a data lake on premises and at the edge of the network continue to expand. However, with the rate data is moving into the cloud, by this time next year usage of Big Data analytics on public cloud services might exceed both in terms of the amount of data being analyzed. It’s not that any single use case for Big Data analytics will supplant another. Rather, all three major use cases are employing machine and deep learning algorithms against more data than ever to deliver more relevant insights faster.
They say data is the new oil. But it’s not until data gets refined that it has any meaningful value to the business. Refining all that data to provide meaningful insights requires the construction of complex data pipelines. Not every MSP needs to have an army of data scientists and software engineers to benefit, though. In many cases, the organizations building these pipelines need to rely more on MSPs to help them secure the networks and protect the volumes of data that are now several orders of magnitude greater than any time in the history of IT. Of course, MSPs that have data science expertise may stand to benefit most. But all things considered, the more data there is to manage and analyze, the better off the entire MSP community tends to be.