Blogs

RedMonk

Skip to content

Enterprising Ubuntu?

In the three years since Ubuntu was unleashed upon an unsuspecting populace, debate as to its viability and marketability within “the enterprise,” whatever that means anymore, has been rampant. Cynics contend that the status quo – where there’s Red Hat and there’s everyone else – is also the shape of things to come. Believers will fire back with shipped CD numbers, all manner of community traction metrics, and isolated but significant ISV engagements.

Predictions aside, there’s little debate that any enterprise strategy is by definition a long term strategy. Unlike consumers, who’ll make Facebook or Twitter worth ten figures more or less overnight (ok, maybe not literally overnight), enterprises move in glacial terms. Geologic, even. [1]

Step one, then, in Ubuntu’s multi-year “enterprise” strategy needed to be “be there in three or four years.” It’s sort of like nature photography; you need to give your subjects time to acclimatize to your presence, and they’ve done that. As UbuntuLive indicated, this will probably go into the books as a check mark. Step two? “Speak their language.” That one, I think, is still a work in progress, although with Mark Shuttleworth discussing otherwise mundane topics like ISV certifications and SI programs and interoperability during his keynote, it seems as the Canonical folks are picking up the unique dialect of larger customers. Step three, in my opinion, is the really tricky one: differentiate. Two announcements made at UbuntuLive are apparently aimed solidly in that direction.

The one getting most of the attention, as far as I can tell, is Landscape. Most of that attention is deserved, I think. Like MySQL and Red Hat before it, Canonical is apparently prepared to augment its software offerings with network delivered services – a trend that most software providers will have to adapt to in the near future, I think. Landscape is, as far as I understand it, essentially a web front end that provides for basic systems administration and reporting for Ubuntu machines.

Is this likely to be differentiating to customers that have employed complicated systems administration tools from the Big 4 or the Systems Management 2.0 players, as my colleague has dubbed them? No. But is it a.) low barrier to entry and b.) new to the Ubuntu world? I’d say that’s a yes on both counts. More importantly, will it open the doors to new customers? Again, I think yes. If it works as advertised, as an example, I am highly likely to apply it to the Ubuntu server (Dapper LTS) powering www.redmonk.com.

The second announcement that I found potentially interesting is the lower profile Personal Packaging Archive service. In its current incarnation, it seems aimed primarily at developers who have projects they’d like to make available to the Ubuntu audience. While interesting, I think it would be far more compelling if PPA offered enterprises the ability to package their internal software in such a fashion that it could be managed by apt. Imagine a world where you could apt-get install any and all of the wild and crufty software that runs your internal servers as easily as Apache, MySQL or PHP? That, I think, would be interesting. Not because that’s not possible already – it obviously is – but because it would lower yet another barrier to entry for would-be enterprise customers. This would require that PPA a.) provide assistance to internal developers that may be unfamiliar with the Debian package format, and b.) local, on-premise services (because some of the software being managed would be considered competitive differentiation). But the investment would be worth it, I think.

What they didn’t announce, of course, was the solution I’ve been asking for – a hard connection between the applications managed and installed by apt/synaptic/etc and a network of qualified experts on said applications. But Mark did hear me out, so it’s not impossible that that idea gets bounced around internally in one form or another. Sooner or later, someone’s going to build it, IMO.

The larger question of whether or not what was announced, or Ubuntu’s continued momentume, will get them onto to the enterprise radar is of course still unresolved. And it should be noted that not everyone is happy about the news.

But there is little question, in my view, that Ubuntu has enterprise ambitions, and if I was Red Hat I’d be paying close attention.

[1] Which is why, despite their lower price sensitivity, they would be a lower priority market for me than SMBs. Bottom up, and all that.

Categories: Linux, Open Source.

  • Pingback: James Governor’s Monkchips » IBM of Course *Does* Support Ubuntu

  • http://www.michaeldolan.com Mike Dolan

    I think the LSB could be the answer. Not as it is today maybe, but in order to have multiple distros “work” in enterprise environments, you need some standardization. The proliferation of distros is fine for enthusiast users – but if you count the % of distros in datacenters, RHEL and SLES will own 99% share. RHEL and SLES also have strong commercial support groups today (RHEL had to ramp it up and imported someone big from Sun to do it well, SLES had to shift from Netware). Ubuntu may have trouble there, but it needs a good support arm for LTS. I have some ideas there… maybe I’ll find them at LW and suggest it… ;-)

  • http://daveshields.wordpress.com dave shields

    Re apt-get, see Apt-Get It or For-get It