Hadoop Summit: Cloudera and Hortonworks who can ride the dust?

One of the biggest hot topics at the Hadoop summit this year is http://www.aliyun.com/zixun/aggregation/13456.html "> Cloudera and Hortonworks the competition between the highly renowned competitors. Now, as the market matures, the rivalry between the two sides has been white-hot, and Cloudera's recent news of a large financial bailout led by Intel has helped. But is this really a life-and-death fight? Or is there enough room in the Hadoop market to hold two heavyweight data giants?

Jeff Kelly, Wikibon analyst, thought about the above issues and exchanged views with Siliconangle founder John Furrier in the Thecube seminar during the first day of the Hadoop summit. Furrier to the point that "in fact, these two companies have serious hostility to each other". He pointed out that Hortonworks recently acquired the Hadoop security professional manufacturer XA secure company is proof. Not surprisingly, Cloudera also responded pointedly, announcing yesterday that it would recruit Gazzang, another business focused on Hadoop security.

"These two stories illustrate the focus of the current Hadoop business," Kelly said. "The straightforward explanation is that it will elevate itself to meet enterprise-level requirements." Vendors such as Hortonworks and Cloudera will actually spend their money and focus on security and data governance challenges – two points will be a critical component if you plan to run Hadoop in a production environment filled with sensitive data in the future. ”

Cloudera and Hortonworks obviously want to make sure that the other side doesn't develop any other competitive advantage, but Kelly thinks the most interesting thing is that the two companies choose a business model that is completely different. Hortonworks has been working to sharpen its distinctive focus by focusing on the fully open source Hadoop core; In contrast, Cloudera has its own Hadoop release, which is supplemented by partially open source components, and the real core is proprietary technology as a functional base.

Four hooves flying, hot hurricane is staged!

At the moment we are at an interesting fork in the road, and these two companies have the opportunity to get huge financial returns from the Hadoop business. As Kelly said, Wikibon recently assessed the size of the big data market and asserted that it would have a total value of $50 billion trillion by 2017-that is, just three years later, the new space will be completely shaped.

"Don't get me wrong, the new big data market is also extremely brutal," Kelly said. "When there is such a huge amount of wealth in the market, Cloudera has a self-financing 900 million dollars and 740 million dollars from Intel, Hortonworks hand in the hands of the full 200 million dollars, the two sides of the rattling will trigger a fierce confrontation between you and me." This is the actual situation in the future, and it is worthy of our interest to pay attention to its actual trend. ”

Cloudera's decision to establish a good partnership with Intel is likely to be a key factor in the situation, especially as the battle with Hortonworks ultimately requires an acquisition to write the ending. "Cloudera and afraid the use of funds to offset Hortonworks's technical advantages," Kelly said. "If competition escalates further, Cloudera's war funds will play a major role, that is to say Hortonworks must find a way to come up with a truly market-attractive outcome and get a new round of financing." ”

In fact, Furrier says, the market has enough room to help Cloudera and Hortonworks to succeed in the Hadoop business, and the evidence is Wikibon's recently published research. He points out that if the factor of the rise of the Internet of Things is taken into account, then two competitors will be able to earn billions of of dollars-at least from the point of view of revenue.

"Eventually, a company will gain a leading edge, and both sides obviously want to get to the top, but in fact, the market space is enough for them to make a pot full of two," Kelly admits. "Not only to spell a champion, secondary seats, then the market will be ranked third of the manufacturers are likely to be very moist." ”

"hadoop--a great opportunity."

But if the Hadoop market does have the ability to accommodate two giants at the same time, it will bring another problem--hortonworks and Cloudera have spent too much time competing with each other, ignoring the most important market itself?

Two companies do have the conditions and need to do some reflection. Kelly revealed some anecdotal information from Wikibon about the latest survey of Hadoop practitioners, and it is clear that this part of the business space will continue to grow and may even pose a serious threat to data warehouse vendors.

"We've just done a series of surveys, and more than 60% of the respondents who are now using the Hadoop program are beginning to give Hadoop responsibility for the workloads that were originally owned by traditional data warehouses or mainframe care," Kelly said. "This is undoubtedly a major shift in the current workload, and the data warehouse vendor's revenue is bound to be affected." ”

This is not to say that every user who uses Hadoop needs to pay for this kind of technology-at least not yet. In fact, most practitioners choose to use the free option, the "self-developed" program or the free HADOP distribution, and only 20% of corporate users currently pay for support services – perhaps to remind us that we are still in the early stages of proof-of-concept. When the Hadoop user completes the proof-of-concept accumulation and begins to move towards a mass-popularization phase, the next problem is to choose Hortonworks and other similar vendors to support their own production workloads-and then we will really see who has the power to command the big data emerging markets.

"The environment we are currently exposed to is still based on proof-of-concept and experimental stages, but when these projects begin to translate into actual production, real development opportunities will come in full," Kelly concludes.