12 Comments

Summary:

This week I’m listening in as HP talks to some of its customers about cloud computing. Today’s webinar was a pretty good overview of how enterprises should think about using the cloud to deliver IT services — and underneath that, why the cloud really isn’t quite […]

black_stretch_logo1This week I’m listening in as HP talks to some of its customers about cloud computing. Today’s webinar was a pretty good overview of how enterprises should think about using the cloud to deliver IT services — and underneath that, why the cloud really isn’t quite ready for enterprise adoption.

David Cannon, who works for HP helping customers use IT to meet business goals, reminded the audience that the cloud right now is a tool that can help large companies by delivering highly scalable compute capacity on demand, but isn’t the underlying capabilities or service that IT needs to deliver to the business or the business’s end customers.

“The cloud makes it possible for you to access resources, so you can don’t have to control IT resources device by device, but with the cloud, the resources are available to everyone. We all have resources, what makes us different is in the area of capabilities we deliver on top of them,” Cannon said.

It was kind of like an enterprise IT pep talk. Which was nice, especially since the rest of the webinar was a bit of a downer for anyone hoping that the availability of on-demand computer resources was going to make their jobs easier. Instead, HP thinks IT’s role becomes one of evaluating the best cloud for a specific job. Things to consider include service-level agreements in various clouds, where a cloud may be located, and several other factors we’ve previously mentioned as well.

The logical solution to these issues is that someone will build a class of enterprise-level clouds, possibly designed to host types of information or applications that require certain service-level agreements, or even to be located in certain geographic regions. Russ Daniels, CTO of HP’s Cloud Services Strategy, agrees, and says HP may even provide a few of these clouds, and will work with customers that will provide even more domain-specific or vertically-focused clouds.

However, Daniels also believes it’s a bit early on to be thinking about portability between clouds, so don’t look for those domain specific clouds yet, unless you’re certain you want to stay with them. But thanks to data exchange standards like XML, he said via chat during the webinar, interoperability is in better shape now.

“The high rate of innovation drives providers to integrated and proprietary architectures — the typical approach in new markets,” Daniels said. “Interestingly providers exploit open source and open standards to build these integrated and proprietary platforms. As the cloud market matures the pendulum will swing back to open standards.”

However, a lack of portability and some vendor lock-in is likely, he thinks, which is yet another reason that enterprises may want to keep their data out of the clouds for a bit longer.

  1. I think it is quick becoming a strategy. Cloud is the back-end for SAS. With the cloud reducing the cost during the build phase SAS can be delivered by small innovative teams. This will accelerate SAS services. Soon the choice will not just be Internal or HP or IBM or India, but will also SAS. Email, laptop backup, personal storage, … all of it is becoming a service in the cloud.

    Share
  2. I agree with the title and not with you (partly) Saul. Cloud is only a tool. Cloud was even available previously but in a different avatar. Why it should be a strategy? To me it is nothing more than an enabler. Enabler to conduct business in a more cost-efficient manner. And I don’t think it will become an enterprise strategy though it will be an important cog in the overall strategic wheel where enterprises will think about providing SAS with reduced cost.

    Share
  3. Nice post. I think HP’s missing the boat my saying it’s not strategic and an overall strategy. I think that moving IT to the cloud can have a dramatic effect on not only the service levels but also the cost structures and really continue to move IT costs into a variable environment. Technologies are really available to deliver solutions in the cloud for many different sized companies. My take is that picking the “simple” services (email, remote backup, video conf, storage) is a cop out and missing the boat completely.

    My take: Move it all to the cloud and leverage the existing Virtual Server solutions, Virtualization of Applications, Published Applications and Desktops, SANs, and deliver on the promise. We’ve been doing this for years based on existing proven technologies from simple Terminal Server desktops to publishing applications out of the cloud. It doesn’t matter what platform, it can be done with extremely high success and SLAs.

    Thanks,
    Bill

    IVDesk.Wordpress.com
    http://www.ivdesk.com

    Share
  4. Considering you say that there’s a lack of portability a paragraph later, I’m not sure what this means: “But thanks to data exchange standards like XML, he said via chat during the webinar, interoperability is in better shape now.”

    Share
  5. Personally, I think that Cloud Computing is BOTH a tool AND a strategy (much like a few others who have replied to this post). When a company is thinking about IT strategies, you can go with the traditional datacenter one or with a cloudcenter one, or even a hybrid model. They are not, however, interchangeable. There are efficiencies that are distinct to each. In my mind, a strategy is a collection of “tools” (not a great word in my mind) and “actions” (decisions on how to use the “tools”). Tools can be hardware, software, networking components, etc. Alone, tools are ineffective and useless, serving only singular purposes. As a “strategy”, these tools are coupled together in the “proper” way to a larger whole, hence the word “toolset”.

    When you architect a strategy or solution, you need to be sure that you choose the correct one. You cannot simply think that if it works within a traditional datacenter, that it will work entirely the same way in a cloudcenter. You must design for the efficiency and strengths of both, depending on which one you choose.

    Put together multiple toolsets in an intelligent manner and then you have a strategy.

    Just my 2 cents.

    Thanks,
    Michael Sheehan
    Technology Evangelist for GoGrid – http://www.gogrid.com

    Share
  6. I think what troubles HP and other hardware vendors is the scalability that distributed computing provides. The price of the bigger and bigger boxes (and thus the margin on them) are increasing exponentially. So building high performance, fault tolerant systems out of commodity PCs is really a big danger for them.

    The article itself is clever. HP is pretty late in trying to jump into the boat. The best it can do is to try to slow the process and hope it can find a place in the new world of cost effective computing.

    The argument of SLAs, vendor-lock-in and availability guarantees are valid IMHO, but I’m pretty sure it will be solved soon.

    BR, David

    Share
  7. [...] 2: Cloud Computing Is a Tool, Not a Strategy [...]

    Share
  8. Tool. Strategy. Incorrect abstraction. Cloud Computing is a business model and technology architecture. It’s an ROI to the penny counters and it’s means to address scalability, disaster recovery and other key IT management issues for IT Management. The question is which definition will reign supreme at the end of the hype cycle?

    Share
  9. [...] Cloud Computing Is a Tool, Not a Strategy by Stacey Higginbotham on GigaOm [...]

    Share
  10. [...] exciting stuff if you ask me. Stacey also provides follow-up perspectives on HP’s presentations here and here. Some interesting takeaways: “HP thinks IT will be delivered as a service by thousands of [...]

    Share

Comments have been disabled for this post