shutterstock_218740615.jpg
shutterstock_218740615 For years, Apache Hadoop appeared ubiquitous within the data-analytics community. Every week, it seemed, more companies voiced their reliance on the open-source framework, ideal for running data applications on large hardware clusters; software vendors offered up proprietary distributions, with claims of improved ease-of-use and faster processing speeds. But has the actual implementation matched the hype? Pivotal, the Big Data-centric spinoff of EMC and VMware, announced back in February that it would stop producing its proprietary Hadoop distribution, perhaps (Venturebeat surmised) in response to a market crowded with similar platforms. Yet despite all the distributions available, a new Gartner report suggests only a minority of companies are actively deploying or experimenting with Hadoop. In a recent Gartner survey, only 26 percent of respondents reported either deploying, piloting, or experimenting with Hadoop; another 18 percent said they plan on investing in the technology at some point within the next two years. Many organizations have downgraded Hadoop’s importance in relation to other software initiatives, and others have claimed the associated costs are too high. "Early Hadoop projects typically involve a small number of users and this no doubt keeps user populations down at this stage of the market,” Nick Heudecker, research director at Gartner, wrote in a brief research note. “Moreover, the Hadoop stack remains unsuitable for simultaneous use by multiple users, also keeping numbers down. Another factor, and possibly the best explanation for the small number of Hadoop users, is the skills shortage." Whether or not Hadoop eventually lives up to its longtime hype, data analytics has become an increasingly vital element of organizational strategy. Tech pros with a background in the various Big Data technologies can easily find themselves in high demand. But which analytics skills pay the best? According to Dice, the top seven, by annual salary, include:

Cassandra: $128,646 MapReduce: $127,315 Cloudera: $126,816 HBase: $126,369 Pig: $124,563 Flume: $123,186 Hadoop: $121,313

Whatever you say about Hadoop, it can still prove lucrative.