cloud abstraction

8/7/2019 Cloud Abstraction

1/14

Pa e 1

Network OSI Model

Data unit Layer Function

Host

layers

Data

7. Application Network process to application

6. Presentation Data representation and encryption

5. Session Interhost communication

Segmen t 4 .Transport End-to-end connections and reliability

Media

layers

Packet 3.Network Path determination andlogical addressing

Frame 2. Data Link Physical addressing

Bit 1. Physical Media, signal and binary transmission

Ha

rdw

are

firmwarehardware

assemblerkernelstorage

memory CPU

Humaninterface

bus

network

Op

era

ting

Sy

ste

m

I/O driversUI drivers

Core components

ServicesSecurity

Userinterface

digital

Interface (API)

Networkvirtualization

Displayvirtualization

Storagevirtualization

Applicationvirtualization

Address space remappingMeta-data

I/O redirection

Hardwarevirtualization

User

Abstraction Layers OSI-style View

Applications / Utilities

Citrix XenAppCitrix XenDesktop

MS Remote Desktop Conn

App-V (fka SoftGrid)Symantec Application Streaming

Citrix Application Virtualization

Citrix XenServerMS Hyper-V

Grid Computing(bus virtualization)

InfiniBandOpenFabrics Alliance

Myrinet

VMware vSphere/ESX


2/14

Pa e 2

Bus

Abstraction Layers Simplified View

Hardware

Network

Operating System

Applications

User

Storage

Each layer can be abstracted/virtualized

BenefitsRapid expansion/deflation

Automated component lifecycleReduced physical footprintReduced physical maintenanceReduced software maintenance

Improved performanceGreater resiliency/redundancy

Instant software updatesReduced cost

Flexibility

Security and privacy

RisksShared database infrastructure

Untrustworthy software installationUnreliable patch managementMixed platforms

Distributed server locationsTransition timeLearning curve

Client security concernsOne infrastructure for diverse applications


3/14

Pa e 3

CloudComputing

Gartner Forecast

Cloud computing services revenue should total $56.3 billion for 2009,representing a 21.3% increase compared to 2008.

The market is expected to explode to $150.1 billion in 2013.

Gartner expects SaaS revenue to total $7.5 billion in 2009, which represents a

17.7% increase over 2008 revenue of $6.4 billion.SaaS demand is expected to continue to surge to a total of more than $14

billion by 2013.


4/14

Pa e 4

CloudComputing

FeaturesAbstraction

(virtualization)

programmaticSelf-serviceInterfaces

scalable

On-demand near-instant scalabilityPay for what you use, use only what you needAbstracted components life-cycle-managedAbstracted components include security layer


5/14

Pa e 5

CloudComputing


6/14

Pa e 6

Corporate Mission Guiding Principles

Corporate Mission StatementCorporate Vision / Superordinate Goals

Production/execution

Management By Objectives (MBOs)

First leg: Goals

Second leg: Action Plan

Third leg: Results

Adaptive Agile/Scrum model

Marketing Plan

Brand Reference Book

Facts Book

Marketing Objectives


7/14

Pa e 7

Corporate Mission Marketing Plan1. Customer Groups

2. Customer Needs3. Technologies

a. Services/offeringsb. Processes/structures

Brand Reference BookFinancial Data

Product DataSales and distribution dataAdv. Sales promo, merch dataMarket data & misc.

Facts BookContinuous Collection1. Review of marketing environment2. Review of detailed market activity3. Review of marketing system

- portfolio planning- 80:20 rule- 7 P's [price, product, promo(adv), placement, people,

physical environment, process, packaging]

Marketing Objectives(matching Corporate Mission Statement)


8/14

Pa e 8

Corporate Mission Production/executionManagement By Objectives (MBOs)

1. Motivation (scoring)2. Communication/coordination3. Clarity of goals

(matching Corporate Mission Statement)Mis-use of MBOsMBO evolution

First leg: Goals1. Specific outcomes, defined requirements2. Scheduling and planning3. Securing support/resources on all fronts

Second leg: Action Plan- Goals only half of what's needed- Simpler is better- Follow-up = Follow-through

- Realistic and flexible

Third leg: Results1. Document and respond2. Recognize high performance

Adaptive Agile/Scrum model


9/14

Pa e 9

Project Management Methodology Investment LifecycleA Short Course in Project Managementhttp://office.microsoft.com/en-us/project/HA102354821033.aspx
http://office.microsoft.com/en-us/project/HA102354821033.aspxhttp://office.microsoft.com/en-us/project/HA102354821033.aspx


10/14

Pa e 10

Project Management Methodology Waterfall (BDUF)

Pros Design is a worthwhile investment Bug fixes in early stages of

development Working out issues in advance

Changing specs cheaper thanchaning systems

Cons Adapts poorly to changing

requirements

Assumes designers can foreseeproblem areas without prototyping Can spill into analysis paralysis Many projects begin with limited or

false requirements


11/14

Pa e 11

Project Management Methodology Agile (eg: Scrum)

Scrum: Developed as a software development tool, adaptiveto maintenance teams, and project/program approach

Roles: Scrum master, Product ownerTeam: Chickens and Pigs

- Daily Scrum- Scrum of scrums- Sprint Planning meeting- Sprint Review meeting- Sprint Retrospective

Agile home ground Low criticality Senior developers Requirements change often Small number of

developers

Culture that thrives onchaos

BDUF home ground High criticality Junior developers Requirements do not

change often Large number of

developers Culture that demands order


12/14

Pa e 12

Project Management Methodology Systems Development Lifecycle(SDLC)

Strengths Control Monitor Large projects Detailed steps Evaluate costs and completion

targets Documentation Well defined user input Ease of maintenance Development and design

standards Tolerates changes in MIS staffing

Weaknesses Increased development time Increased development cost Systems must be defined up front Rigidity Hard to estimate costs, project

overruns

User input is sometimes limited


13/14

Pa e 13

Project Management Methodology SDLC Phases


14/14

Pa e 14

Project Management Methodology Dynamic Systems DevelopmentMethod (DSDM)

Agile home ground Low criticality Senior developers Requirements change often Small number of

developers

Culture that thrives onchaos

BDUF home ground High criticality Junior developers Requirements do not

change often Large number of

developers Culture that demands order

cloud abstraction

Documents