Articulating the value of Cloud computing

1,965 views
1,870 views

Published on

This was the keynote presentation given at the Microsoft Cloud Architect Forum in London on Sept 25th 2009. It explains the value of cloud computing from a variety of perspectives and shows how the model fits with other technology models that people are familiar with. It then explains some of the main ways that it makes sense to use cloud computing resources and highlights some gotchas.

Published in: Technology, Business
0 Comments
3 Likes
Statistics
Notes
  • Be the first to comment

No Downloads
Views
Total views
1,965
On SlideShare
0
From Embeds
0
Number of Embeds
2
Actions
Shares
0
Downloads
0
Comments
0
Likes
3
Embeds 0
No embeds

No notes for slide

Articulating the value of Cloud computing

  1. 1. Articulating the business value of Cloud Computing Microsoft Cloud Architect Forum Neil Ward-Dutton Research Director
  2. 2. Before we can talk about value, we have to be clear what we’re talking about! <ul><li>This is a problem that’s not unique to Cloud Computing, of course </li></ul><ul><li>Hype, confusion, bandwagon-jumping always abound with hot new technology trends </li></ul><ul><ul><li>Cloud 2.0, anyone? </li></ul></ul><ul><li>We need to clearly define Cloud Computing scope and capabilities before we can be clear about its business value </li></ul>© MWD Advisors 2009 www.mwdadvisors.com
  3. 3. Cloud Computing: evolution, not revolution Web services / SOA Timeshare computing ASP, web hosting Data centre outsourcing Billed access to remote computing resources Distributed, standards-based computing / integration Access to software services over the Internet SLAs, managed operations © MWD Advisors 2009 www.mwdadvisors.com
  4. 4. What is Cloud Computing? Architectural, economic, strategic elements Computing and storage resources providing an application platform © MWD Advisors 2009 www.mwdadvisors.com Elastic resource capacity Utility pricing Economic element: Pay-as-you-go, pay-as-you-grow, no capex Abstracted resources Management automation Self-service provisioning Architectural element: Simple, abstract environment for development Third-party ownership Managed operations Strategic element: Focus on what makes you better, leave the rest to someone else
  5. 5. Another way to look at Clouds: Who does what? Raw computing, storage resources Datacenter environment/resources Break/fix, basic monitoring Software platform mgt Capacity provisioning, scaling Usage metering, management Customer-owned / managed datacenter = customer’s responsibility = service provider’s responsibility = not a focus Billing © MWD Advisors 2009 www.mwdadvisors.com Service provider on-premise managed infrastructure Raw computing, storage resources Datacenter environment/resources Break/fix, basic monitoring Software platform mgt Capacity provisioning, scaling Usage metering, management Billing Raw computing, storage resources Datacenter environment/resources Break/fix, basic monitoring Software platform mgt Capacity provisioning, scaling Hosted, managed infrastructure Usage metering, management Billing “ Pure” Cloud Computing proposition Raw computing, storage resources Datacenter environment/resources Break/fix, basic monitoring Software platform mgt Capacity provisioning, scaling Usage metering, management Billing
  6. 6. Cloud is a model of utility service consumption, not technology ownership Open, publicly accessible Restricted access Provider-owned, managed Customer-owned, managed Private Cloud Public Cloud Adaptive Computing Infrastructure (ACI), “legacy” infrastructure © MWD Advisors 2009 www.mwdadvisors.com “ Private Cloud” is about restricted, secure access to Cloud services – not customer ownership We use the term “Adaptive Computing Infrastructure” to refer to “Cloudy” infrastructure that you can buy
  7. 7. Overlapping value: Cloud Computing and its cousins Public, Private Clouds Elastic resource capacity Utility pricing Abstracted resources Management automation Self-service provisioning Third-party ownership Managed operations Outsourced Data Centre Adaptive Computing Infrastructure © MWD Advisors 2009 www.mwdadvisors.com
  8. 8. Why you should consider Cloud Computing and SaaS as separate Outsourced business processes In-sourced business processes Business processes “ Finished” business applications Application development and deployment platform (developer / administrator languages, tools) Infrastructure (processing, storage, networking resources) Cloud-based applications (SaaS) “ Traditional” applications Cloud-based platforms (PaaS) Existing development tools, platforms Cloud-based infrastructure (IaaS) Managed infrastructure Adaptive Computing infrastructure (ACI), legacy infrastructure Provider-owned, controlled Customer-owned, controlled © MWD Advisors 2009 www.mwdadvisors.com Both exist in a world of many choices
  9. 9. The benefits look interesting – what about the flipside? Economic elements Utility pricing, elastic capacity Architectural elements Virtualisation, automation, provisioning Strategic elements Third-party ownership, management Benefits Drastically reduce or eliminate Capex on IT infrastructure Opex scales in line with actual need and usage Rely on providers’ ability to generate economies of scale re: cost of hardware, power, cooling, etc Reduce need for some management skills © MWD Advisors 2009 www.mwdadvisors.com Development platform is abstracted from physical infrastructure – no need to worry about cross-platform testing, etc Potential drawbacks Skills Will new platforms allow me to reuse existing development, architecture skills? Security How can I be sure that my data is secured and protected, in accordance with regulations? Integration How can I integrate remote systems with others (either remote or on-premise) with good enough QoS? Lock-in If I deploy applications to a Cloud, how can I be sure I’ll be able to move from one to another if I need to?
  10. 10. What do you want on your Cloud? Different Clouds, different tradeoffs Choice/flexibility High Low Built-in platform richness High Low Amazon AWS [1] Google App Engine (GAE) [2] Force.com [3] Joyent Accelerator [1] AWS requires a particular virtual machine format (AMI), but very few other restrictions [2] GAE provides an application framework of sorts, but only supports development in Python or Java [3] Force.com provides a very rich application framework but also imposes a very strict application programming model [4] ASP ties you to a .NET based programming model, but it provides a lot of richness and also spans public and private resource deployments Microsoft Azure Services Platform (ASP) [4] Rackspace Cloud © MWD Advisors 2009 www.mwdadvisors.com
  11. 11. Cloud Computing: into the mainstream <ul><li>Numerous large-scale examples of online services built on public Cloud resources </li></ul><ul><ul><li>Twitter, etc </li></ul></ul><ul><li>Initial public Cloud take-up by online startups: now mainstream interest from enterprises </li></ul><ul><ul><li>Washington Post, Harvard Medical School, etc </li></ul></ul><ul><li>Mainstream enterprise infrastructure software vendors defining Cloud Computing strategies, products, services </li></ul><ul><ul><li>Microsoft, IBM, TIBCO, Oracle, Citrix, ... </li></ul></ul><ul><ul><li>For many it’s simply a “stick it on AWS” strategy, or an Adaptive Computing Infrastructure (ACI) strategy </li></ul></ul>© MWD Advisors 2009 www.mwdadvisors.com
  12. 12. 3 entry points for Enterprise Cloud investment: Aligned with key Cloud elements Economic elements Utility pricing, elastic capacity Architectural elements Virtualisation, automation, provisioning Strategic elements Third-party ownership, management <ul><li>Non-uniform workloads </li></ul><ul><li>Where the expense of acquiring infrastructure is difficult to justify </li></ul><ul><li>Application development, test, staging </li></ul><ul><li>Large dataset batch analytics </li></ul><ul><li>Online promotions </li></ul><ul><li>Rapid time-to-market </li></ul><ul><li>Where a ready-to-use platform is highly compelling </li></ul><ul><li>Service/application prototyping </li></ul><ul><li>High business demand for new application functionality </li></ul>“ Cloud by stealth” Where an existing service provider implements own Cloud infrastructure to gain economic advantages © MWD Advisors 2009 www.mwdadvisors.com
  13. 13. Selling Cloud to the business: highlighting business cases <ul><li>DO acknowledge challenges - potential benefits are striking, but potential drawbacks might make people nervous </li></ul><ul><li>DON’T try and sell “Cloud Computing” as a general concept </li></ul><ul><li>DO look for particular applications where the business case is clear and where perceived risks can be accepted </li></ul><ul><li>DON’T be tempted to “just try it out” for live applications without testing potential legal implications first </li></ul>© MWD Advisors 2009 www.mwdadvisors.com

×