4 Comments

Summary:

On this week’s Structure Show, OpenStack Foundation executive director Jonathan Bryce and COO Mark Collier make their case for enterprise OpenStack adoption.

The Structure Show 2014

OpenStack promises a lot: Robust clouds that can support both specialized private and massive public workloads but without fear of vendor lockin. And that is a lot to promise. On this week’s show, we’ve got not one but two heavy-hitting proponents, Jonathan Bryce and COO Mark Collier, telling us why, exactly, big companies should check out the new Icehouse release of the open-source cloud framework (available Thursday!)

OpenStack Foundation COO Mark Collier (L) and Executive Director Jonathan Bryce (R)

OpenStack Foundation COO Mark Collier (L) and Executive Director Jonathan Bryce (R)

To satisfy one key demand, Icehouse will enable rolling upgrades from the previous without shutting down workloads. Such upgrades in the past, frankly, were a pain, indeed, some called them backbreaking , so if this works as planned, it’s a big deal.Also new, support for federated identity, a feature requested by CERN, the European nuclear laboratory /OpenStack poster child.

More than a dozen companies from traditional IT purveyors like HP and Cisco to startups like Cloudscaling and Mirantis, are backing OpenStack as the infrastructure to give customers choice in cloud deployment. If Acme Paint & Glass signs on to OpenStack A and isn’t satisfied it can, in theory,  move to OpenStack B without too much hassle. But the vendors all say they’re innovating atop OpenStack and differentiation often means incompatibilities. But Collier said the Foundation will be enforcing good OpenStack citizenship going forward. Go too far afield in your OpenStack playbook and BOOM! you lose your OpenStack label. (My words, not his.)

At the Structure show  in June  there will be much more on what’s moving and shaking in the cloud ecosystem, so stay tuned.

But first Derrick Harris weighs in on Microsoft’s pitch to make Azure the big data platform for the masses and we discuss how cloud vendors including Amazon Web Services will address end-of-life processes for creaky old infrastructure. Also, why Twitter bought Gnip. Or is it Nip? Whatever, it’s a good one so listen up!

 

SHOW NOTES

Hosts: Barb Darrow and Derrick Harris

Download This Episode 

Subscribe in iTunes

The Structure Show RSS Feed

PREVIOUS EPISODES:

Enterprise cloud adoptoin: Amazon is big, natch, but watch out for vSoup

Cloudera CEO breaks down Intel investment

Don’t like your cloud price? Wait a second

Pro tip: how to learn to love the cloud (and not worry when it shoots you through the head)

How Declara brings together cloud, AI and social networkst to help us learn better

 

 

  1. Barb, there was lots of examples of increasing momentum for OpenStack at the Red Hat Summit this week. Executives from numerous major vendors — and some early-adopter end users — shared insights about why they believe this open-source project has such widespread support.

    Perhaps we’ll learn more details about other case studies at the upcoming OpenStack Summit in Atlanta. The most compelling case for OpenStack adoption seems to come from experienced CIOs who have grown weary of the proprietary hardware and software constraints of the past.

    Share
    1. jonathan and mark said to expect new big users at the OpenStack Summit including some banks. i think this is what they need to show…. big companies doing more than POCs and not all this OpenStack vendors selling (or bartering) withother OpenStack vendors.

      Did you see/hear from any big new OpenStack customers at red hat that haven’t been done to death?

      Share
    2. In what way does it do that at all in the real world? It’s a theory that doesn’t play out in any actual implementation. At the end of the day you need hardware on prem to run workloads. The operational overhead of managing that hardware lifecycle is specifically what businesses want to escape. Adding *more* hardware vendors, or white box builds (insanity in an enterprise) just exacerbates the issue exponentially.

      On the software side the story is similar. Service and support is where the money comes from. Businesses spend a ton on “free” Linux and end up with just as much fragmentation and operational cost as with any proprietary OS. Only initial capex is (somewhat) reduced.

      Private cloud *in general* has extremely questionable value for any enterprise beyond just smarter orchestration of traditional virtualization. The inherent vale of “cloud” is getting you out of the infrastructure build business and putting you in the service broker business. A science project like open stack ironically puts you *deeper in* to the build business.

      If you look at the real adoption of open stack it’s the whose who of “we have no story against AWS and Azure”. No way something as fragmented and poorly defined as open stack will be a magic bullet for the likes of lagging providers like HP, IBM and Cisco.

      Share
  2. Jamie Costello Thursday, April 17, 2014

    It would be great to hear success stories of Open Stack used in production without an army of engineers or consultants to support the Open Stack environment. I was at the “Open Stack at mega scale” meetup in the Bay area recently. I asked the panelists (eBay, Yahoo and Symantec), and each had 15+ engineers just on getting Open Stack up and running, and on day to day operations. Thats headcount you will not need if you went with public clouds.

    Share

Comments have been disabled for this post