Cloud.  Big data.  These are elusive terms to which we pin a myriad of meanings.  But Mike Olson, CEO of Cloudera, has a handle on the cloud, big data and how they relate to the enterprise.  Cloudera’s been a driving force behind Hadoop’s entrance into the enterprise sector, and Olson’s company is thriving on innovation and execution, prepped for tomorrow’s data management needs.

In this week’s Profile Snapshot series, Olson discusses how Cloudera uses the cloud, what every Hadoop startup should know, and how he spent his last vacation (even CEOs need to take a break sometimes!).

How do you use the cloud?

We’re an anomaly in Silicon Valley these days: With “cloud” in our name and cloud applications available all over the place, we own a huge farm of rack-mounted servers housed in a colo. We use it for development and testing of Apache Hadoop and related projects. We can’t ship our software without making certain that it works at scale, so we buy a whole lot of computers.

More constructively, though, we’re big users of Zendesk for our customer support infrastructure, we use both Box and Dropbox for file sharing, our global sales team collaborates on Salesforce and we use Marketo to feed the field. Atlassian is the source of all truth at Cloudera — we couldn’t operate without our wiki and JIRA ticketing systems. And, of course, we hit AWS for all sorts of on-demand stuff.

What are three things every Hadoop startup should know?

One: There’s great money to be made building apps that run on Hadoop, so please do that! We need that innovation to drive business adoption of the platform. Two: Hiring is brutal. Your best bet is to hire great people and teach them Hadoop. Deep Hadoop internals skills are awfully thin on the ground. Three: Cloudera loves you, and our Cloudera Connect program is how we show it, so check it out!

What’s in store for Cloudera in 2012?

Scale!

You’ve leapt at least 2 years ahead of the competition with your vision of Hadoop–what does “balancing the future” mean to you?

It’s pretty clear to me that we’re past the point where Hadoop is interesting standalone technology.  A lot of the earliest internet companies that adopted it used it as a massive, and isolated, data storage and compute platform. In the more traditional businesses rolling it out today, though, it gets installed alongside relational databases, existing enterprise applications and all kinds of critical infrastructure.

Hadoop has to play nicely in data centers that look like that. We’ve changed our view — which was, frankly, pretty parochial and Hadoop-centric when we got started almost four years ago — so that we look at the business problems that confront an organization, and the existing investments that the organization has made. We concentrate more on use cases and on integration. The data center is undergoing a pretty radical transformation now, with new data streaming into elastic and cloud infrastructure, ubiquitous virtualization and new platforms like Hadoop for tough problems. We need to help customers make that transition.

Let’s Get Personal!

What’s your personal philosophy?
.
Get close to interesting people.

What was your last vacation?

Over the New Year’s holiday, I turned off email for three days and hung out with my wife. It was wonderful.

What’s on your iPad?

My daughter’s fingerprints, thanks primarily to Angry Birds. I just finished Steve Jobs’ biography on iBooks (seemed the right way to read that), and I am addicted to Instapaper.