• Usable for business, Optimized for efficiency due to high cost
CLIEN T-SERVER
• PCs and servers for distributed compute, storage, GUI, etc.
• Optimized for agility due to low cost and ease of use.
CLOUD
• Large DCs, commodity HW, scale‐out, multiple devices • Order of magnitude better efficiency
and agility
COM PUTIN G PARADIGM DISRUPTION S
TECHN OLOGY BUSIN ESS CASE
M ore than just a tech shift… com pelling business case
Cloud computing is a pay-per-use model for enabling available, convenient, on-demand network access to a shared pool of configurable computing resources
(e.g., networks, servers, storage, applications, services) that can be rapidly provisioned and released with minimal management effort or service provider
interaction (NIST – National Institute of Standards & Technology).
Presenter
Presentation Notes
Self service model – administration model can be managed by the procuring entity Shared pool of resources – the offering should utilize share resources (servers, storage, network, power) so as to maximize economies of scale Scalable and elastic – the system should have the ability to scale as needed, instantly Metered by use – the system should only charge the using entity for the resources consumed Broad network access – the system should be capable of deploying services to a variety of platforms and device types across public, or widely available private networks. Key points: Not just Web apps or someone else’s datacenter. From a competitive standpoint important to broaden the concept of cloud across devices. Script: First of all, “the cloud” is short for cloud computing. As with many topics that blend technical subjects with business hype there are a number of definitions. Most of the time people think of large datacenters that someone else runs on your behalf or Web applications that you can access from browsers. We tend to think of it as something more, it’s the collection of devices (servers, PCs, mobile devices) using the network to pool resources and work together. You typically see this in large datacenters today; where servers seamlessly work together to easily absorb more traffic and to react when something breaks without people having to intervene. The sum ends up being greater than the parts. So, it’s about datacenters and devices and applications working better together. So what? Why should I care? In general you can use the cloud in three separate ways: Pushing things to the cloud, Using the cloud as a resource when needed (or all the time) and sharing or hosting your data or applications with others once you’ve put it into the cloud (one you directly control or that could be run for you by someone else.)
Application runs on-premises
Buy my own hardware, and manage my own data
center
Application runs at a hoster
Pay someone to host my application using hardware
that I specify
Application runs using cloud platform
Application runs on-premises
Application runs at a hoster
Public Cloud Business Case
Less Control over data
Traditional
Storage
Servers
Networking
O/S
Middleware
Virtualization
Data
Applications
Runtime
You
man
age
Infrastructure(as a Service)
Storage
Servers
Networking
Middleware
Virtualization
Data
Applications
Runtime
Managed by provider
You
man
age
Platform(as a Service)
Managed by provider
You
man
age
Storage
Servers
Networking
O/S
Middleware
Virtualization
Applications
Runtime
Data
Software(as a Service)
Managed by providerStorage
Servers
Networking
O/S
Middleware
Virtualization
Applications
Runtime
Data
O/S
Operations Operations Operations Operations
Presenter
Presentation Notes
Slide Objectives: Explain the differences and relationship between IaaS, PaaS, and SaaS in more detail. Speaking Points: Here’s another way to look at the cloud services taxonomy and how this taxonomy maps to the components in an IT infrastructure. Packaged Software With packaged software a customer would be responsible for managing the entire stack – ranging from the network connectivity to the applications. IaaS With Infrastructure as a Service, the lower levels of the stack are managed by a vendor. Some of these components can be provided by traditional hosters – in fact most of them have moved to having a virtualized offering. Very few actually provide an OS The customer is still responsible for managing the OS through the Applications. For the developer, an obvious benefit with IaaS is that it frees the developer from many concerns when provisioning physical or virtual machines. This was one of the earliest and primary use cases for Amazon Web Services Elastic Cloud Compute (EC2). Developers were able to readily provision virtual machines (AMIs) on EC2, develop and test solutions and, often, run the results ‘in production’. The only requirement was a credit card to pay for the services. PaaS With Platform as a Service, everything from the network connectivity through the runtime is provided and managed by the platform vendor. The Windows Azure Platform best fits in this category today. In fact because we don’t provide access to the underlying virtualization or operating system today, we’re often referred to as not providing IaaS. PaaS offerings further reduce the developer burden by additionally supporting the platform runtime and related application services. With PaaS, the developer can, almost immediately, begin creating the business logic for an application. Potentially, the increases in productivity are considerable and, because the hardware and operational aspects of the cloud platform are also managed by the cloud platform provider, applications can quickly be taken from an idea to reality very quickly. SaaS Finally, with SaaS, a vendor provides the application and abstracts you from all of the underlying components.
Studies show that a 100,000‐server datacenter can have 80% lower total cost of ownership (TCO) compared to a 1,000‐server datacenter.
Typical DCs PUE : 4Best in class enterprise DC PUE : 2
1.000
1.200
1.400
1.600
1.800
2.000
2.200
2006 2008 2010 2012
Microsoft Average PUE Targets for New DCs
PUE
Modern Cloud DCs
Presenter
Presentation Notes
PUE: Power Usage Effectiveness The ratio of total energy consumption (servers + cooling) to 'useful' energy consumption (servers only).
Average UsageCompute
Tim e
Usage
Compute
Tim e
Average
Inactivity
Period
Compute
Tim e
Average Usage
Compute
Tim e
Average Usage
Presenter
Presentation Notes
Not 100% of apps; majority. Predictable bursting – busy in day, idle at night. To meet the market response to successful services. Unknown acceptance to application. HPC At the platform level… If you need new functionality or systems or want to add extra compute or storage to existing systems then there are a few workload patterns that work best for the cloud. Predictable bursting Services with micro seasonality trends Peaks due to periodic increased demand IT complexity and wasted capacity Growing fast Successful services needs to grow/scale Keeping up w/growth is big IT challenge Complex lead time for deployment Unpredictable bursting Unexpected/unplanned peak in demand Sudden spike impacts performance Can’t over provision for extreme cases On and off On and off workloads (e.g. batch job) Over provisioned capacity is wasted Time to market can be cumbersome
Discussion –Potential Cloud Usage PatternsPattern Areas of interest
1 Transference :Taking existing on-prem ise app and m oving it to the cloud.
• Com pany Em ail, W eb Site.• New business m odels –SaaS.• Alternative software deploym ent
m odels.
2 Scale & M ulti-Tenancy:Abilityto handle web load without full capital investm ent from day 1.
• Testing the m arket.
3 Burst Com pute:Ability to handle additional com pute on an as needed basis.
4 Elastic Storage:Creating an application where storage can grow exponentially.
• Businessintelligence.• Genom ics.• Load testing on large databases.
5 Inter-Organization Com m unications:Application that has the ability to com m unicatebetween organizations using the cloud as the pre-defined hub.
• Distributed developm ent.• Collaborative R&D.
Standards and Choice in Developm ent
Azure™ Platform
HTTP
RSS
SAM L
OpenID
W S-TrustSOAP
Atom Pub
W S-M an
REST
W S-Eventing
Standards based interoperability
Presenter
Presentation Notes
Key Points: Script: Businesses are moving to the cloud. Analysts predict that the vast majority of large businesses will be using some type of cloud services in the next few years and that many will drastically be cutting the IT assets that they run and maintain in favor of pushing it to others to run for them. If you’re looking at a new IT project or you’re facing some new business challenges, it’s worth taking a look to see if the cloud can help you. This all probably sounds good from a theoretical standpoint but whose actually doing this today?
Presenter
Presentation Notes
• Shift to “Cloud First” policy. Each agency will identify three “must move” services within three months, and move one of those services to the cloud within 12 month and the remaining two within 18 months.
• The three-part strategy on cloud technology will revolve around using commercial cloud technologies where feasible, launching private government clouds, and utilizing regional clouds with state and local governments where appropriate.
• Reduce number of Federal data centers by at least 800 by 2015
Value
Core Platforms (DB, App Server)
Middleware (EAI, EII, ETL)
Customization
Operating System (Windows, Linux, Mainframe)Virtual Machine, Management Tools
Dev
elop
men
t Too
ls
Hardware
Value
Core Platforms (DB, App Server)
Middleware (EAI, EII, ETL)
Software Services and Innovation
Operating System (Windows, Linux, Mainframe)Virtual Machine, Management Tools