Keynote Speech: How Much Of Your Future Will Be In The Cloud? Strategies For Embracing Cloud Computing Services, James Staten, Principal Analyst, Forrester
Cloud computing has shifted from being a question of “if” to one of “when” and “where” in your IT future and portfolio. Is it best to stick with SaaS, or should you be deploying new services directly to the public clouds like Amazon EC2 or Windows Azure? What applications are candidates for the cloud, and which should remain in-house? And for how long? This session will explore the enterprise uses of cloud computing thus far and synthesize the thinking across Forrester on this issue to present you with a road map and a strategy for embracing the cloud that benefits both your business and the IT function. Cloud can be a catalyst for the IT-to-BT transition so long as you harness it effectively.
Session attendees can expect to learn:
- How to tell a true cloud solution and its relative maturity from simple cloud washing.
- The truth behind the economics of cloud computing.
- The best places to start and strategies to build your own path to cloud efficiency.
Prior to the conference, James wrote a positioning/discussion piece, which is published on ZDNet. From what I saw on Twitter, the most controversial idea was the “Pay per use or metered consumption” requirement to be consider cloud computing.
James opens with “Cloud computing isn’t an if, it’s a when and a how”. He says it won’t change your entire world (Nick Carr), nor is it complete hype (Ellison). But, you will use it, situationally.
Definition: A standardized IT capability (services, software or infrastructure) delivered in a pay-per-use and self-service way. On the pay-per-use, you should be able to go to zero. That’s the economic power, according to Staten. On self-service, the request is immediately processed, in an automated manner. [This eliminates all the “managed service” cloud players].
James says enterprises need to take a portfolio approach to IT infrastructure and applications. Make best cloud/not cloud decisions.
On private clouds, James offers a reality check. Cloud infrastructure requires the following. Is your enterprise there?
- Standardized operating procedures
- Fully automated deployment and management
- Self-service access for deployers
- Business units sharing the same infrastructure
[James says the 4th is the hardest. I would (respectfully) disagree. Cloud computing or not, the business units should not care, nor have a say, in the underlying technology infrastructure, or the placement of their applications within it. If business people are worrying about IT infrastructure and deployment, then who is worrying about the actual business?]
Now, James is talking about virtualization maturity. He has a 4-stage journey: acclimation, strategic consolidation, optimization and “cloud”. At “cloud” you have automation, self-service, optimization etc. James cautions the audience on virtual machine sprawl. While you would save on physical resources, the management of virtual machine sprawl can be counterproductive.
In a portfolio approach, James has dedicated infrastructure, virtualized infrastructure and private cloud. The scale goes from custom to common, fixed to transient, owned to metered.
In a slide transition, James adds a huge Public Cloud circle. Don’t forget/discount the public cloud. If public cloud isn’t part of your portfolio, you are missing out on economic benefit of going to zero.
James gives some examples of cloud cases on Forrester site: Associated Press (time to market gain), PathWork Analytics (run heavy data analysis, then go to zero), Data.gov (cost reduction story).
Now James is telling the Animoto story (scale on EC2), and other business disrupters leveraging the cloud, such as bizo, Good Data and Flight Caster.
Move from Capex hard commitment to Opex. James places public cloud computing in a broad outsourcing family, along with managed hosting and SaaS. Difference for cloud is the ability to go to zero.
Organizations will need a decision tree to place right applications in right deployment to get best economics for business. Two factors are workload management and GRC.
Next 90 days:
- Built a cloud leverage team
— Invite your most innovative thinkers
— Use hybrid architectures to optimize deployments
— Build a Cloud management model
Longer term
- Start in pockets
— test and development
— green-field applications
— throw spaghetti (public cloud)
- Create a cloud migration road map (beyond dev/test and green-field)
Or, fall for “cloud is hype line” at your own peril.
Related posts:
- Forrester to Enterprise Architects: Use Cloud Phenomenon to Boost Role as Business Advisor
- Forrester: Interactive Data Protection Heat Map
- James Urquhart: It’s not (just) the data destination, it’s also the journey
- James Urquhart: Understanding Infrastructure 2.0
- James Urquhart: Cloud computing’s green paradox
{ 1 trackback }
{ 0 comments… add one now }