I attended IBM’s Pulse conference earlier this month, which has traditionally been focused around the former Tivoli division. It’s is a strange mix of software for sysadmins and software that can do things like manage city water supplies (the GreenMonk, Tom, is all about the latter). Last April, IBM renamed it Cloud and Smarter Infrastructure — even the short form, C&SI, doesn’t exactly flow off the tongue.
While I don’t write about every conference as a rule, this one was interesting for a variety of reasons. One is the increasing focus on cloud to the point that they called Pulse “The Premier Cloud Conference” this year, including a gigantic banner across the keynote stage. Somehow I doubt that went over terribly well with today’s customer and user base of traditional IT and asset-management people, as my colleague James also noted while at the show. IBM has a very interesting and solid developer story here, but much like VMware did with its developer story in the pre-Pivotal days at a VMworld full of IT admins, IBM’s presenting to the wrong audience. I found the story they told, including live coding on a stage at an IBM business conference, really compelling for developers. The problem was that it’s an aspirational message at this point about their audience and their conference rather than one that’s a good fit for the conference as it stands today.
The second reason Pulse was interesting is because of IBM’s actually doing with the cloud now. Since the days of SmartCloud, which was widely viewed as a subpar cloud offering, IBM’s gained a lot more credibility through involvement with OpenStack, its purchase of SoftLayer, and significant investment in the Cloud Foundry PaaS as well.
IBM’s always been about selling business value, not technologies, so its move up the stack from IaaS to PaaS is a natural extension of that focus. This shift is consistent with other moves IBM has recently made, such as progressively selling off more and more of its commodity hardware businesses — first ThinkPad and more recently xSeries. IBM sees much greater success when it sells business results, not commodities, and PaaS is a better fit than IaaS for that reason.
Cloud Foundry has gained a great deal of traction in the past couple of years, so IBM’s choice of it as a platform makes a great deal of sense. As that happened, we’ve seen Cloud Foundry gain broader governance than its previous leadership solely by Pivotal and earlier by VMware, again in line with what I’d expect from IBM. Choosing an open-source PaaS is in-line with IBM’s selection of OpenStack and, more than a decade ago, Linux, so I see this move as a good one that’s clearly been thought through and that is consistent with IBM’s philosophy and behavior.
The bigger picture
I think we’ll continue to see IBM’s software divisions reorient around the “IBM as a Service” concept that it pushed at Pulse. In the context of BlueMix, that will mean an ongoing addition of buildpacks based on its existing software portfolio as it’s able to repackage them as services.
I further see this as a revival of IBM’s commitment to key open-source projects as far back as Linux and Eclipse nearly 15 years ago. The extent to which IBM’s bought in to OpenStack and Cloud Foundry, and to a lesser extent Chef, is significant — it’s no matter of a fly-by-night patch series, it’s a major effort that’s headlined as critical to IBM’s future to its own customer base. There’s an excellent opportunity for IBM to continue extending this commitment to open source, as we’ve seen to an extent with Hadoop, Worklight (via PhoneGap), and across other areas like social, WebSphere, and the rest of information management. If it cares about developer traction, and it appears to do so, open source is a clear route to success in the form of an olive branch.
Disclosure: IBM, VMware, Pivotal, the Eclipse Foundation, and Chef are clients, as are numerous OpenStack and Hadoop vendors. The OpenStack Foundation is not.
No Comments