Imagine your company spent more than $100 million developing a product. Now imagine that a competitor came along and cloned your product and distributed a near-perfect replica of it. Not good, right? If you're Apple, you spend years and tens of millions of dollars fighting it, determined to be the one and only source of your product.
If you're Red Hat, however, you embrace it—as Red Hat CEO Jim Whitehurst told ReadWrite in an interview.
For years the enterprise data center was defined by expensive hardware running varieties of the Unix operating system. Over time, both Windows and Linux chewed into Unix's market share, with Red Hat winning the bulk of the Linux spoils. The key to victory? Both Windows and Linux offered low-cost, high-value alternatives to Unix's sky-high pricing.
With Unix cowering in a corner, one would think that the battle would shift to Linux versus Windows. The reality, however, is somewhat different.
As Whitehurst tells it, Red Hat "certainly competes" with Microsoft, but "generally those IT decisions are made at the architecture level before you get into a specific Linux versus Windows bake-off." Today enterprise architecture tends to be Linux-based, while 10 years ago it was Windows, which means that more often than not, Red Hat Enterprise Linux is baked into enterprise IT decisions.
"Going forward with new workloads, they are heavily Linux-based," notes Whitehurst. As such, Whitehurst doesn't "worry about Microsoft long-term, because it's Red Hat and VMware that are defining future data center strategy."
Taking On VMware
Ah, yes, VMware. Sun Microsystems, in its day the leading Unix vendor but now swallowed up by Oracle, once provided Red Hat with a handy villain to target. Today data-center software maker VMware is Red Hat's Enemy Number One.
The reason is simple: No other company more closely matches Red Hat's ambitions, albeit with a very different approach. As Whitehurst emphasizes, "When you start thinking about where the future of the data center is going, VMware has a similar view to ours, but they're doing it with a proprietary innovation model and we're open."
How open? So open that not only is Red Hat fighting VMware with its own open-source products, but it's also embracing clones like CentOS.
While open source is increasingly established within the technology world, few understand its implications for an open-source software business. In the case of Red Hat, it develops the popular Red Hat Enterprise Linux (RHEL) operating system. But because Linux is a community-developed OS, Red Hat must release all of its Linux code to others. (Instead of charging for a software license per se, Red Hat has customers pay for a subscription that covers services and support.)
This paves the way for an organization like CentOS to develop a "a Linux distribution derived from ... a prominent North American Enterprise Linux vendor" which "aims to be 100% binary compatible" with that Linux vendor.
It's the imitator that dare not speak its name, but everyone knows CentOS is a like-for-like Red Hat clone. How can this possibly be good for Red Hat?
Embracing The Parasite
While some like Microsoft have threatened Red Hat with the specter of even greater competition from CentOS, Whitehurst argues that CentOS "plays a very valuable role in our ecosystem." How? By ensuring that Red Hat remains the Linux default:
CentOS is one of the reasons that the RHEL ecosystem is the default. It helps to give us an ubiquity that RHEL might otherwise not have if we forced everyone to pay to use Linux. So, in a micro sense we lose some revenue, but in a broader sense, CentOS plays a very valuable role in helping to make Red Hat the de facto Linux.
But couldn't another Linux vendor like SuSE or Canonical, the primary backer of Ubuntu, undercut Red Hat with an equally free OS? If $0 is the magic price point, other Linux vendors can easily match that, right?
Whitehurst responds: "SuSE often comes in at a lower price point than RHEL, but most people would prefer to have a common code base like RHEL plus CentOS than a cheaper but always fee-based enterprise SuSE."
In other words, only Red Hat can offer the industry's leading Linux server OS and also offer—albeit indirectly—that same product for free.
Microsoft has tacitly acknowledged a similar phenomenon: While the company spends heavily to fight piracy, founder Bill Gates noted in 1998 that illegal copies of its Windows operating system in China helped seed demand for the paid version.
While I'm sure Red Hat's salesforce doesn't love competing with its copycat, the reality is that sales are almost certainly helped in accounts that only want RHEL for production servers and can shave costs by using CentOS for development and test servers. CentOS, in other words, gives Red Hat a lot of pricing leverage, without having to lower its prices.
Arguably one critical area that CentOS hasn't helped Red Hat is with developers. While developers want the latest and greatest technology, Red Hat's bread-and-butter audience over the years has been operations departments, which want stable and predictable software. (Read: boring.) CentOS, by cloning RHEL's slow-and-steady approach to Linux development, is ill-suited to attracting developers.
So Red Hat is trying something different, dubbed Red Hat Software Collections. Collections includes "a collection of refreshed and supported web/dynamic languages and databases for Red Hat Enterprise Linux." Basically, Collections give developers a more fast-moving development track within slower-moving RHEL. Or, as Whitehurst tells it, "Collections is Red Hat's way of embracing developers while keeping its appeal for operations."
It will be interesting to see how this plays out. Red Hat has a long way to go in its goal to define the open data center, but with its embrace of CentOS to give it licensing leverage and of Collections to give it developer credibility, Red Hat is on the right track.