While the walls of my home are mostly monopolized by pictures of my kids, there is a picture of me with a couple dozen twenty-somethings. It is a picture of Red Hat’s engineering team from 1999, right around the time we took the company public. I had the privilege of leading that team as it built the products that still form the core of Red Hat’s $8 billion business.
While Red Hat was building the basic software that runs many of the largest companies in the world (enterprises like Cox, McKesson and Union Pacific), Intel was changing how computers were being deployed. Very large computers from companies like Sun and IBM were being replaced by large numbers of computers built around Intel’s microprocessors.
Fifty Sun machines would be swapped out for a thousand Dell machines. Each one of those new machines is almost identical to the computer you have in your home. Making them at large scale made them much, much cheaper than the large, complex systems they replaced – so much cheaper that buying 20 times the number of machines saved companies a lot of money.
All of those computers needed software, and most of them got either Red Hat’s Linux offerings or Microsoft’s Windows products. The Internet was growing rapidly, and data centers were expanding.
All of this encouraged me and a couple of other folks from the picture to leave Red Hat and start rPath, a new company trying to create new ways to manage the thousands of servers that organizations were deploying. We all had a shared background in the open-source technologies Linux is a poster child for and strong relationships with the companies that were deploying large numbers of Linux boxes. Our history and skills made large-scale Linux deployments the most obvious place to go look for customers.
While Linux was a market for us, our idea for managing the systems was built around version control. We wanted to enable system administrators to put version numbers on every piece of software in their businesses: operating systems, applications, databases and custom software. We feel like one of the keys to managing thousands of computers is to be able to define exactly what version of each software is on each one, have reliable ways to automatically deploy those versions to every machine, and automatically check those systems to make sure they have the software they are supposed to have and that it’s deployed properly.
Fast forward a few years, and rPath has customers using our technologies to manage tens of thousands of Linux systems in their data centers. While we are thrilled with our customer list, those same customers made it clear that Linux comprised only part of their systems-problem difficulties. Their Microsoft Windows deployments are large and complicated, and they started asking us how our version-control technologies could help those machines.
Those conversations set off some interesting discussions. Our customers helped rPath focus on what our core values are: version control and system automation. As we realized that the benefits we provide are not Linux or Windows values, the decision to add support for Windows to our product line became simple. There was certainly a lot of work to do to bring the product to market, but a year later we’re excited to be working with our first Windows customers, bringing them the agility and reliability benefits that system version control delivers.
rPath is looking forward to helping our customers conquer the complexity of their growing data centers through our core technology, whatever software they choose to use to make their business a success.
Erik Troan, is co-founder and chief technology officer for rPath, the service factory for on-demand IT. Learn more about rPath and follow rPath on Twitter at @rpath and on Facebook. Contact Erik at ewt@rpath.com.
Get the latest news alerts: Follow LTW at Twitter.
