Startup Rocana wants to be everything Splunk is and more, bringing big data to DevOps
An example Rocana dashboard.
When big data first hit the tech scene, it was all many people would talk about. A plethora of BI and analytics tools hit the market using big data technology, but certain aspects of the enterprise were left behind—namely operations.
Years back, Omer Trajman, Eric Sammer, and Don Brown met while working together at Cloudera. They were early executives and field employees charged with helping customers understand Hadoop and how to use it to solve business problems. They did this, possibly, hundreds of times and, while they were able to help many business better understand Hadoop, they always seemed to hit a snag when it came to getting it into production.
The problem was that the operations team, the folks in the data center, didn’t know how to run it or weren’t able to. So, they started solving the operations problem as well. They screwed it up plenty of times, Trajman said, but they also got it right many times as well.
“We started realizing that we could actually productize it—there was a need in the market for a new product for IT operations, and that we could solve this problem better than what was available using concepts from big data,” Trajman said.
Thus, in early 2014, Rocana was born. The company essentially applies big data, advanced analytics, and visualizations to DevOps—giving admins better transparency into their data centers and applications and helping them solve problems more quickly.
Rocana currently employs about 40 people throughout North America and has raised just under $20 million in two rounds of fundraising. Investors include General Catalyst, Google Ventures, Toba Capital, Brian Stevens, and Paul Sagan.
The Rocana team.
Using Rocana Ops, admins get a better look at components like web servers, app servers, databases, firewalls, routers, device metrics, and cloud services. Rocana calls this “augmented IT operations,” and their tool is available through a web browser.
Trajman said there are three main concepts behind the product:
1. Collect everything
“You don’t really know, ahead of time what data is going to be important until it’s too late,” Trajman said.
Executives don’t realize that it’s possible to collect everything, Trajman said, and they have been conditioned to believe that everything must be filtered. However, that’s simply not the case with today’s technology.
2. Eliminate silos
The second concept behind Rocana is the belief that data sources don’t have to be separated. For example, by collecting and presenting data together, they’ll be able to see a relationship between anomalies in networking data and application data.
“What we’re looking for is not just anomalies in the data,” Trajman said. “We’re not just looking for deviations in patterns. We’re looking for areas where those have been correlated.”
In the past, Trajman said, users could name servers and draw images of their architecture. Now, applications span data centers and third party services. Additionally, cloud and virtualization introduce a new level of ephemerality. Infrastructure is now more complex than it’s ever been. Rocana crunches the data to point users to the areas where they should be focusing their energy and where issues are arising.
This approach to big data in operations is precisely the reason that Steve Goodman and the team at Toba Capital agreed to invest in Rocana.
“What was compelling about what Omer and his team were doing was not necessarily the technology that they’d developed, it was the way they were thinking about the problem,” Goodman said.
Rocana isn’t stepping onto an empty battlefield. Log analytics company Splunk has been providing similar services for more than a decade and is a strong contender. However, Rocana claims it is offering everything Splunk is and more. Trajman said Rocana is the “first time that companies can see everything.”
Over the past few months, a heated dispute has formed between the two companies, beginning with a Rocana blog post by David Spark outlining four key differences between the two products. Splunk responded with a formal cease and desist letter regarding claims made about their technology in the post.
Needless to say, the two companies probably won’t be inviting each other over for the holidays.
Monitoring technology and log analytics are nothing new, but they have suffered from siloing in the past. Tools like Rocana could help operations admins get a better look at problems and take care of them sooner.
Looking forward, Trajman said the company is working on connecting the dots faster. Right now, they can show users where a current problem exists, but they want to be able to predict potential problems based on patterns in the data and eliminate the threat of “cascading failures” that come with cloud scalability.
As the company continues along its path, Goodman said one of the biggest challenges in front of them is keeping up with the demands of the enterprise and the quickened pace of deployments, without over extending themselves.
“What Rocana really needs to be careful of is that they don’t try to do too much too quickly,” Goodman said.