Red Hat CEO Whitehurst on VMware, OpenStack and CentOS

Red Hat CEO Whitehurst on VMware, OpenStack and CentOS

Summary: 'Open source gives us brand permission to enter a ton of categories,' said Red Hat CEO Jim Whitehurst.

SHARE:

Red Hat has been on an acquisition run with the aim of building an open source stack that leads the hybrid data center. The primary rival: VMware.

How the Red Hat vs. VMware duel plays out remains to be seen, but Jim Whitehurst, CEO of Red Hat, argued that open source will ultimately define the enterprise IT architectures of the future.

I caught up with Whitehurst to talk cloud, open source and Red Hat's master plan.

Here's a look at some of the key themes from my conversation with Whitehurst:

JimWhitehurst

Recent acquisitions and the big picture. Red Hat recently acquired OpenStack integrator eNovance and Inktank, a software defined storage company known for Ceph, an enterprise platform. Whitehurst said that Red Hat "is looking to create an open source stack for infrastructure and platform as a service." Red Hat makes acquisitions not based on intellectual property as much as people and involvement with innovative open source projects like Ceph, an object and block storage platform.

Ceph is a way to play scale-out storage. Whitehurst argued that storage will move to open source because of cloud computing and new analytic workloads. OpenStack will also mean that technologies such as Ceph will be adopted over time.

New workloads. The game for Red Hat and its storage and cloud platform plans really revolves around capturing new workloads in the enterprise. "We tried the migrate and replace approach with Unix to Linux. A much easier sell is using open source technologies like JBoss for new workloads and then migrate. Cloud deployments are always new workloads," said Whitehurst, who added that Red Hat's Gluster storage technology is going for new unstructured data workloads.

Docker and Linux containers: Red Hat opens up on the issues

Red Hat Enterprise Linux 7 and incentives to upgrade. When asked why customers would migrate to RHEL 7, Whitehurst said the question missed the point. He doesn't care when they upgrade. "One of the most important parts of Red Hat's business model is we don't care if you upgrade or not. If you're a support subscriber it doesn't matter," said Whitehurst. "It doesn't economically matter to us if customers upgrade or not."

RHEL 7, which supports containers and packs a bevy of other features, will have an upgrade cycle that resembles previous releases. Customers will run existing workloads on whatever RHEL they're running today. New workloads will go to the latest version because it's built for cloud deployments both public and private.

Red Hat moves forward with new OpenStack release

CentOS 7 is on its way

A big step forward in business Linux: Red Hat Enterprise Linux 7 arrives

OpenStack. Whitehurst's enthusiasm for OpenStack hasn't waned at all. He notes that Red Hat is the biggest contributor to OpenStack and the company's goal is to make it stable, secure and enterprise class. Fragmentation isn't an issue, said Whitehurst. "A lot of competitors sell against OpenStack and say it's fragmented. It's no where near as fragmented as Linux."

Red Hat's approach to OpenStack is unique because it's selling support not hardware or consulting services. There's no incentive to make OpenStack less open by adding proprietary technologies. "We have a different value proposition then IBM or HP," said Whitehurst.

Defining Red Hat. Whitehurst chafed when I asked whether Red Hat should be seen as a Linux or cloud company. Red Hat is an open source company and that's the compass that leads every move it makes. "The way our DNA works is that we look at the most innovative open source projects that are applicable to the enterprise," he said.

That DNA has led Red Hat to the cloud, but it has led to Linux as an OS to middleware to platform as a service and virtualization. In the future, open source will lead Red Hat to networking.

"Open source gives us brand permission to enter a ton of categories," said Whitehurst. "We're looking at all technologies most valuable to the enterprise. At some point networking will be a logical thing for us and we're very involved with the Open Daylight project. We do open source and make it a production system no matter what the category."

CentOS. CentOS has generated some buzz because it's a lighter version of RHEL. One argument is that CentOS could be interesting to the enterprise because companies could hire their own Linux admins for support over Red Hat. Whitehurst said that thinking is misguided. "CentOS is a derivative of RHEL that works for people that want a stable release without support," said Whitehurst. "If they don't see the value of our model then at least I'd rather have them on something similar to RHEL."

The other item to note with CentOS is that Red Hat hired the top contributors to the project. "We don't view CentOS as a competitor. It's almost a complement," said Whitehurst, who noted that enterprises going for CentOS wouldn't be Red Hat customers anyway.

Special Feature

The Art Of The Hybrid Cloud

The Art Of The Hybrid Cloud

Cloud computing is insatiably gobbling up more of the backend services that power businesses. But, some companies have apps with privacy, security, and regulatory demands that preclude the cloud. Here's how to find the right mix of public cloud and private cloud.

So what's the value for Red Hat? Whitehurst said CentOS gives developers a faster cadence to do new innovative things relative to RHEL, which is all about stability. "Red Hat is great for production, but there are so many new innovations with OpenStack," said Whitehurst. "There was nothing in the Red Hat family to support that. CentOS is moving faster and fills a gap in our portfolio of offerings."

In other words, CentOS gives the developer community something to play with and Red Hat can contribute and learn at a faster clip.

Red Hat vs. VMware. VMware is clearly Red Hat's No. 1 rival, said Whitehurst. "The software-defined vision is the same, but the two models are very different," said Whitehurst. "Our model is open source and theirs is licensing IP."

Whitehurst acknowledged that VMware has a strong position because it is the private cloud's defining architecture due to its virtualization share. In the long run, however, OpenStack will garner a lot of attention.

"The defining architecture for the hybrid cloud will be OpenStack or vCloud," said Whitehurst. Red Hat will do well if "we have a position in the right communities." It's unclear whether the hybrid cloud will be about managing virtual instances (VMware) or containers (OpenStack). "We don’t have a strong architectural position, but if we're involved with the communities where the innovation is happening then the value is in working with us," said Whitehurst.

One key issue for VMware will be all the vendors in the OpenStack ecosystem. The common bond between them: "No vendor in the ecosystem wants to create the next Microsoft. That's one key reason why there's so much support for OpenStack," said Whitehurst. "Customers don't want to be locked in either. Even VMware customers talk to Red Hat because they want an open alternative."

Topics: Cloud, Enterprise Software, Linux, Open Source, VMware

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.

Talkback

11 comments
Log in or register to join the discussion
  • Re with open source your not locked in...

    What a load rubbish regarding the statement with Openstack (or any other open source software) that customers are not locked in, of course they are.

    It does not matter whether the software is open or 'closed', once a customer makes a major investment in the software they are always locked in as they have made a financial commitment and to the training of staff.

    When will people be 'open' and honest!
    pjc158
    • You have no clue

      "Lock in," is short for "vendor lock in," meaning that if you want support you are forever locked in to only going to that one vendor that gave you technology in the first place. With open source you will never have that problem, because the vendor who gave you the technology also gave away that code that makes it. That means you or any company you contract can pick up that source code and start providing you with support. Open source eliminates vendor lock in.
      T1Oracle
      • Exactly.

        I would only add the worst part of Vendor lock in, IMO: When the vendor doubles or quadruples the price at new version time.
        Case in point: MSSQL version 2012: The pricing model changed so that a customer with a 2 socket server was forced to pay 4x as much for the 2012 version, compared to the previous version for the same server. And, unlike Windows, MS gave no option to continue buying the older version once the new version was available.

        This caused one of our partner companies to switch to PostgreSQL, because the MSSQL 2012 pricing pushed the price of their document storage product beyond the price range of their customer base.
        anothercanuck
        • But, your partner company wasn't locked in ...

          Rabid Howler Monkey
    • I can get out..

      If I am on a none OSS propriety custom system, it's hard to get my stuff out if the maker wishes to hold onto it.. with OSS, I can use standard protocols and tools to get my data. it's like office formats before people reverse engineered them.. locked users to ms office.
      frankieh
    • Yeah, no.

      OpenStack is vendor independent. Obviously once some entity starts down one path, they are, in a practical sense, "locked in". But if you are locked into what some exclusive provider gives you, you basically have no recourse if they decide to go in a direction you do not like. With open source technologies, you can swap out providers/vendors/suppliers and heck, even the software itself, and lose little-to-nothing on your overall investment.

      People in IT know and understand. People in upper management....not always.
      wantoosevin
  • Microsoft merit its technology

    On just being a really good alternative to open source that a lot of customers love. What companies like Redhat hate about Microsoft is they have a complete end to end stack that is very integrated and harmonious. Which company, both open source and proprietary has the following:

    Cloud OS - Azure
    Private Cloud OS - Windows Server, Hyper-V
    Enterprise Client - Windows
    Consumer Client - Windows
    Embedded Client - Windows Embedded
    Mobile OS - Windows Phone
    Developer Tools - Visual Studio
    Productivity - SharePoint, Exchange, Lync, Yammer, Office 365, Office desktop
    Tablet/Mobile device - Surface/Lumia

    With such a complete end to end scenario, which provides a balance between the open hardware and yes, a proprietary ecosystem, you get tools and services that ready and supported.

    Competitors to Microsoft needs to move away from the thinking of Microsoft, being the Microsoft of 1996 with NT 4 and Office 97. Microsoft is an excellent brand and even if its proprietary, who give a f..., fact is, the platform, stack and ecosystem is top notch. They have what competitors like Redhat is just attempting to put together and its proven. Windows is proven, Office is proven, Azure is proven, Windows Server is proven.
    adacosta38
    • "Just jealous?"

      Let's call this one "MS standard defense number two" ("all corporations are equally corrupt" is number one); I think it's a straw man.

      The major objection to MS over the past 25 years or so has been its talent for positioning itself as the only realistic choice in the markets in which it competes; Bill Gates was a genius at doing just that; Steve Ballmer was rather clumsy at it, but he still tried (that's really what the patent wars and the War on Google are all about, as the Browser Wars were before them). The antitrust cases of the late 1990s and early 2000s have handicapped MS' ability to overtly "cut off the air supply" of competing firms and products, but efforts to maintain dominance by all possible means continued throughout the Ballmer years (but I don't have time to substantiate that at the moment). Satya Nadella has an opportunity to abandon traditional practices and compete on the merits, but it is very unclear at this point that he intends to do so.

      Personally, I have zero objections to MS or any other vendor offering end-to-end solutions and doubt that any other serious MS critic who posts here does. What I object to are the efforts to make it as difficult as possible to use a personal computer (other possibly than those made by Apple) without paying MS for the privilege. This includes the (largely successful) efforts to browbeat Android vendors into making royalty payments on undisclosed MS-held patents. This includes the long running effort to discredit/villify Richard Stallman and the free software movement. It includes the now-failed efforts to turn the WWW into a IE-only platform and to prevent outsiders from copying MS' file formats and protocols. And it includes the effort to encourage cross-licensing of software patents between large vendors (making it easier to force upstarts and freelancers out of the market; which is what I think the purpose is).

      If people want to use MS' products, then I have no objections to their doing so; but MS and its "partners" should respect the consumers' right (not merely privilege) to make other choices. I still hope to see an MS that competes solely on the merits, but wonder if it will still be offering anything I might want to buy when that happens.
      John L. Ries
    • Why the MSFT advertising?

      The article isn't even about them, and VMware is mentioned chiefly because it is the single biggest competitor in the industry. Both RHAT and MSFT are merely challengers in that space.

      I have worked extensively with both Linux based and Microsoft based ecosystems and can attest to Microsoft's ability to integrate within their technology stack and the strength of their developer tools both in stability and ease of use. They have also improved the overall security and stability overall since the dark and awful days of winXP and 2003 server and prior. There are also promising signs that they are taking seriously the need to be solutions oriented rather than the insular lock in machine they have been traditionally.

      On the downside for MSFT is that said legacy of vendor lock in is still entrenched. Non Microsoft technology is still far too much of a second class citizen in this ecosystem. Vendor support is adequate but still inferior or their competition, and community support is very weak relative to typical major open source projects. Also as "proven" as their cloud and services and such are, they are still not as mature as some of what VMWare offers.

      Red Hat on the other hand has complementary strengths. Their support is stellar compared to MSFT. Their products being Free software centric are widely supported by a larger community of developers and users from multiple vendors. If outside support is a priority Red Hat would wipe the floor with MSFT. They are also active and cooperative contributors in open standards and platforms and projects throughout the stack you detailed. As such third party integration is addressed in a much better way. With openstack for example you can choose from a number of hypervisors, storage backends and networking options in the solution regardless of vendor. That means vendor lockin is not a major issue

      RHAT conversely is weak where MSFT is strong. They are not the owners of all the code they offer in their solutions and so cannot offer the same level of seamlessness...the same integration exercise applies to their own offerings and third party alike in many cases. And there is no open source IDE that matches Visual Studio in us ability.

      I think it is personal choice in the end, but an open model like RHAT offers is a better move for long term support and scalability especially for large enterprises (how many outfits with the scale of Google, Facebook, twitter and the like use the club see MSFT stack after all?). But if you are smaller or unwilling or unable to self support or invest in vendor support then MSFT is indeed worth examining, and with a new CEO and strategy that appears more correct than before the option may get more appealing over time.
      Mark Hayden
  • I read on SJVN's blog that Red Hat was in discussion

    with the Scientific Linux developers to potentially bring them under the umbrella as it did with CentOS. This might open up a desktop category and more.
    Rabid Howler Monkey
    • Addition: Meant to add a workstation category as well

      Rabid Howler Monkey