Cloudera’s Mike Olson: “This isn’t about knocking over old guys and stealing their wallets"

This article was originally published on Technology.Info.
As part of our continuing strategy for growth, ITProPortal has joined forces with Technology.Info to help us bring you the very best coverage we possibly can.

Hadoop distributor Cloudera’s co-founder and chief strategy officer, Mike Olson, talks candidly to Technology.info about the open source framework’s growing maturity, that investment from Intel and whether rumours of the death of data warehousing have been greatly exaggerated.

Q: Forrester analyst Mike Gualtieri recently described Hadoop’s momentum as “unstoppable”. Where, in your opinion, does the technology sit today in the minds of

enterprise customers

?

A: Hadoop’s come a long, long way.

Cloudera’

s been in business for just about six years now. We were incorporated in June 2008 and, at that time, ‘Big Data’ was not a meme. The idea just hadn’t occurred to most people and Hadoop was pretty rudimentary back then. In the ensuing six years, Cloudera has led the community in driving new capabilities: we’ve added NoSQL engines to our offering, like Hbase and Accumulo; we’ve invented, and given away as open source, high-performance SQL query capabilities, in the form of Impala; we’ve integrated other open source projects, like Solar, to deliver search capacity; we’ve embraced third-party tools, like the Spark engine from Databricks, for complex analytic and stream processing. We give our customers lots of ways to get at data, as well as capabilities around security, data lineage, governance, back-up, disaster recovery - enterprise-grade features that were missing from platform that Yahoo and Google first used. That’s a long way for me to say that, while the market is still emerging, it’s no longer fair to see Hadoop as an early-stage technology. We’re seeing it adopted across every single vertical, we’re seeing customers roll out large-scale deployments, and pay seven, even eight-figure sums, on an annual basis, to do just that. Broad adoption may be only just now accelerating but the momentum is tremendous.

Q: Back in April, Cloudera announced a huge $900 million round of investment, $740 million of which came from Intel. What can you tell us about that?

A: Well first, I’d say that it’s not just Intel making big investments. Look at the considerable investment that IBM’s making in big data. Look at the investment in Pivotal, the VMware/EMC spin-off, made by General Electric. Large enterprise infrastructure vendors everywhere are ploughing enormous money and staff into the big data opportunity built around Hadoop.

Talking about our relationship with

Intel

specifically, I understand why people hear a sum like $900 million dollars and imputed market capitalisation of $4.1 billion and they stop listening after that, because those are just breathtaking numbers. What I want them to hear is that we did not do this deal for the money.

The relationship with Intel does three important things for Cloudera. First, it aligns our two companies behind a single platform that we can bring to market together, rather than battling for customers with our own Hadoop distributions. Intel had its own distro and it was pretty good. It has a lot of features that we, frankly, coveted. By aligning behind a single distribution, we can marry the best of both.

Second thing is, we’re a platform company that relies on channel relationships and partners to help us deliver a full stack of value to our customers. Intel has decades-long partnerships across every sector of hardware and software. It’s deeply respected and genuinely liked by all the big players in the market. So by working with Intel, we create a much larger indirect channel and amplify our selling effort significantly.

The third and most significant point for me, is that the silicon we run on is going to change dramatically over the next few years - the balance between memory, disk, compute and networking will be radically redrawn. Intel knows what the future looks like. And we’re in a position to ensure that the open-source ecosystem takes full advantage of all that silicon goodness, that it adapts to these ratio changes in data centre infrastructure. We’ll be able to deliver a much better, faster, more secure platform to the market broadly, before anyone else will.

Now Intel believes the opportunity is enormous and wanted to participate meaningfully, so as a condition of the large commercial relationship, wanted to take a substantial equity stake in Cloudera. We weren’t looking to dilute our shareholders to that extent. We weren’t looking for $900 million in cash - it’s a crazy amount of money for a software company to raise at this stage. But what we did want was that commercial relationship with Intel and so we negotiated an investment that we think makes sense to everybody.

Q: Over at your competitor Hortonworks, executives tell us that Cloudera’s message is that the traditional data warehouse is dead. Can you clarify for us where you actually stand on this issue?

A: Everybody wants to know if Cloudera is after Teradata. I say to them, “Look: the opportunity here isn’t to knock over old guys and steal their wallets.” The opportunity is to monetise vast amounts of data using tools that weren’t previously available. Now

Hortonworks

and Teradata have a joint go-to-market strategy and Hortonworks CEO [Rob Beardon] is on record saying that Hadoop cannot be real-time, that it will not do interactive queries in the way data warehouses do. Cloudera’s point of view is that that’s ridiculous: not only will it, but right now, today, with our platform, it does. The

data centre of the future

will be different. The enterprise data hub will absolutely be a core component of every large enterprise’s data management strategy. Data warehousing will evolve to do more things, better - but an enormous amount of what happens in the data warehouse today is moving out. It’s moving out for flexibility, for cost and because most of the innovation is going to happen in open source, not in single-vendor proprietary technologies. We’re not interested in preserving existing markets. We’re interested in innovating so that customers can do more with their data. So our mission really isn’t to replace the data warehouse. It’s to make data management and data infrastructure better.

See Cloudera's Chief Architect, Doug Cutting keynote as part of the

Big Data Evolution Summit

at IP EXPO Europe, 8 - 9 October 2014, London ExCel.

Topics