redefining cloud computing again linthicum with bonus

Redefining Cloud Computing…Again

David S. LinthicumDavid@bluemountainlabs.comTwitter: @DavidLinthicum

Abstract:

In this session we’ll look at what cloud computing is, and what it should be. The way it’s defined today, and how should be defined

tomorrow. We’ll focus on the emerging patterns of cloud computing, including changing

business value and new patterns of cloud computing technology that are emerging.

Where To Find Out More

Bonus Content: Cloud architecture best practices.

Slides available:

Where to find me.

• @DavidLinthicum• david@bluemountainlabs.com• www.bluemountainlabs.com • www.davidlinthicum.com• http://www.linkedin.com/in/davidlinthicum

• Grab a card after the presentation.

Geek Level for this Presentation

Systems Programmer andDungeon Master

Your Grandparents that are Still Using AOL Dialup

The Problem

• Current ways of looking at cloud computing are too broad and too limiting at the same time. – No good categorization

of sub-patterns.– No definition around

where things are going, more so than where they have been.

Why I’m speaking out…

• There seems to be an increasing overuse of cloud computing concepts as saviors for all past IT mistakes.

• Pushing cloud computing as the way to solve all, or even most, computing problems reveals those who make such statements as less than credible.

So, what is cloud computing?

Architecture?Technology?Use cases?

What will it be soon?

What is the future?

1940

1950

1960

1970

1980

1990

2000

Rise of Timesharing

2010

Grids

Distributed Computing

Datacenters Emerge

Rise of the PC

Rise of “The Cloud”

Rise of Client/Server

Rise of the Web

2010

2012

2014

2016

2018

2020

2022

IT On-Demand

2024

DistributedService Sharing

Hardware/Software/Infrastructure On-Demand

Rise of “Big Data”

Rise of “IT In-a-Box”

Rise of “Commodity Data Services”

The “BigMigration” Begins Rise of Shared

Enterprise Business Services

Rise of “HomeClouds”

Cloud Computing is Becoming Systemic

• Buzzword “cloud computing” is absorbed into computing.

• Focus on fit and function, and not the hype.

• Security moves to “centralized trust” models.

• Centralized data becomes a key strategic advantage.

• Mobile devices become more powerful, but thin.

• The rise of the “composite cloud.”

• Focus on the reorganization and outplacement of data.

• Focus on PaaS, and service companies that are good at PaaS.

• Focus on centralized trust, including moving to identity management models.

• SOA patterns and technology find new value in the cloud.

• Continued focus on mobile computing.

• Home clouds (e.g., iCloud) create a new track of application and appliance development.

• Rise of the “cloud aggregator.”

From here to there…

NIST defines cloud computing as a set of characteristics, delivery models, and deployment models

On-demand self-service

Ubiquitous network access

Resource pooling

Rapid elasticity

Pay per use

5 Characteristics

Software as a Service (SaaS)

Platform as a Service (PaaS)

Infrastructure as a Service (IaaS)

3 Delivery Models

Private Cloud

Community Cloud

Public Cloud

Hybrid Cloud

4 Deployment Models

Delivery Models/Technology Morphing

• Software as a Service (SaaS) – Applications as a Service– Utilities as a Service– Connected and Disconnected

• Platform as a Service (PaaS)– Design as a Service– Process as a Service– Testing as a Service

• Infrastructure as a Service (IaaS)– Database as a Service – Management as a Service– Middleware as a Service– Integration as a Service – Information as a Service

…and more.

Infrastructure-as-a-Service

Secu

rity-as-a

-Serv

ice

Storage-as-a-Service

Inte

gra

tion

-as-a

-Serv

ice

Database-as-a-Service

Information-as-a-Service

Process-as-a-Service

Pla

tform

-as-a

-Serv

ice

Application-as-a-Service

Management/Governance-as-a-Service

Testing-as-a-Service

New Technology Stack Emerging

Understanding the Emerging Patterns of Architecture

Private CommunityPublic

SOA Shared Services Shared Services Shared Services

Hybrid

Security Private Trust Central Trust Tiered Trust

Data Central Distributed Tiered

Provisioning Self Self/Auto Managed

Tenancy Multi/Distributed Multi Managed

Reference Architecture

Data

Data Services/Abstraction

Reference Architecture

Data

Data Services/Abstraction

Transactional Services

Reference Architecture

Data

Data Services/Abstraction

Transactional Services

Process Management (BPMS) Composites/PortalsRules Management

Reference Architecture

Security & Identity

Managem

ent & Service G

overnance

Data

Data Services/Abstraction

Transactional Services

Process Management (BPMS) Composites/Portals

Monitoring and Management

Rules Management

Types of Cloud Architecture Patterns Emerging

• Migration to a Cloud– Business Systems– Infrastructure

• Building on a Cloud (PaaS)• Building a Hybrid Cloud • Building a Cloud

– IaaS– SaaS– PaaS

Complexity

So, It’s Something Like This

Cloud Computing

Public, Private, Hybrid

Applications

Visual (UI Based)

Non-Visual (Services)

Infrastructure

Storage

Block

Object

Compute Network Management

Self Provisioning

Auto Provisioning

Database

Structured

Unstructured

Security

Encryption

In Flight

At Rest

Identity

Application Development

Design

UI

Program Flow

Process

Programming

Code Control

Revision Mgmt.

Testing

Component

Regression

Deployment

Provisioning

Management

Middleware

So, what is cloud computing?

• The concept of cloud computing is about the ability for organizations to stop solving all IT problems by themselves.

• Cloud computing certainly about sharing resources, such as storage and compute services, but it really should be more about sharing solutions and pushing risk out of the business. – Unfortunately, the focus today is on the tactical side of

sharing computing resources. – Should be more about the impact to a business rather than

how the technology is defined.

Q&A

We are leading the conversation with thought leadership and cutting edge services.

The shift to Cloud

#1 BEST SELLING BOOK ON CLOUD COMPUTING

#1 MOST READ BLOG ON CLOUD COMPUTING

#1 MOST LISTENED TO PODCAST ON CLOUD COMPUTING WITH OVER 10,000 LISTENERS

David S. LinthicumBlue Mountain Labsdavid@bluemountainlabs.com

How to Get Cloud Architecture and Design Right the First Time

Abstract

So, what are the proper ways to design, build, and leverage cloud computing systems? What are the steps to success? What are the emerging best practices? How does SOA fit in? In this session we’ll answer these questions, and more, covering the right and wrong ways to leverage, design, and build cloud-based systems and infrastructure. Going beyond the hype, this includes advice from those currently in the trenches who make cloud computing work for the Global 2000 and government. The audience is anyone who will soon fight to make cloud computing work for their clients, employers, and/or investors, and need to knock it out of the park the first time.

Where To Find Out More

Bonus Content: How to architect a hybrid cloud computing solution.

Slides available:

Where to find me.

• @DavidLinthicum• david@bluemountainlabs.com• www.bluemountainlabs.com • www.davidlinthicum.com• http://www.linkedin.com/in/davidlinthicum

• Grab a card after the presentation.

Geek Level for this Presentation

Systems Programmer andDungeon Master

Your Grandparents that are Still Using AOL Dialup

Why You Should Care

Most Cloud-Based Systems are Lacking Architecture

The results:• Inefficient utilization of resources. • Resource saturation.• Lack of elasticity and scalability. • Lack of security and governance. • Frequent outages. • Bad or no tenant management. • Other very bad things.

Common Mistakes

• Not understanding how to make architectures scale.

• Not dealing with tenant management issues.• Not understanding that security is systemic. • Not understanding the proper use of

services.• Tossing technology at the problem.• Listening to the wrong people. • Managing-by-magazine.

The Multitenant Thing• No two multitenant

architectures are alike.• Needs to be effective in

the context of the problem.

• Needs to leverage best architecture practices.

• Needs to practice isolation and protection.“With a multitenant architecture, a software application is

designed to virtually partition its data and configuration so that each client organization works with a customized virtual application instance.”

Source: Wikipedia

The ABCs Multitenancy

Source: Lori MacVittie

ABCs of Multitenancy

Source: Steve Bobrowski Amazon Machine Images

Virtualization is Not Always a Requirement

Types of Cloud Architecture Patterns Emerging• Migration to a Cloud

– Business Systems– Infrastructure

• Building on a Cloud (PaaS)• Building a Hybrid Cloud • Building a Cloud

– IaaS– SaaS– PaaS

Complexity

General Architecture Patterns to Follow

Enterprise Vulnerabilities Assess your Risks

1. Governance2. Regulatory Compliance3. Security & Identity Management4. Business Continuity5. Process and Services6. Data Management7. System Integration8. Resource Skills & Knowledge9. Application Readiness10. Network Readiness

Assess your vulnerabilities and mitigate your risks.

COMPLEXITY OF MITIGATING RISK

RIS

K T

O O

RG

AN

IZA

TIO

N

12

5

34

6

7

8

9

10

Architecture Roadmap Where to beginBuild a solid foundation:

Determine business drivers

Understand: data, services, business processes and system integration points

Design a “cloud-friendly” enterprise architecture roadmap that leverages Services Oriented Architecture (SOA)

Determine an integration strategy for internal and external systems

Outline a migration path for legacy systems to the new architecture

Leverage SOA to build a sturdy and agile foundation for the Cloud

“as-is”

“to be”

deploy

General Rules and Guidelines

• Focus on the primitives. – The best clouds are sets of low level services that can

be configured.• Data services• Transaction services• Utility services

• Leverage distributed components that are centrally controlled.

• Build for tenants, not users. • Don’t lean too much on virtualization. • Security and governance are systemic.

Use Case: General Business System Migration to the Cloud

The General Idea

• Break the system down to it’s component parts, understanding each part.

• Rethink the database, it’s typically a mess.• Rethink the system as sets of services/API, watch the

granularity. – Data services, transaction services, utility services

• Create a security and governance plan. • Consider the user interfaces.• Consider device interfaces.• Migrate only the components to the cloud that will

provide the best value.

Good New/Bad News

• Good News:– Most cloud platforms provide the tenant

management features for you.– Most cloud platforms manage the resources

for you.– The cloud platforms are getting better.

• Bad News:– Security is still your problem.– Application and service design is still your

problem.– Testing can be a hassle.

Reference Architecture

Security & Identity

Managem

ent & Service G

overnance

Data

Data Services/Abstraction

Transactional Services

Process Management (BPMS) Composites/Portals

Monitoring and Management

Rules Management

Create the InformationModel

OntologiesOntologies

Understand Ontologies

Understand the Data

DataDictionary &Metadata

DataDictionary &Metadata

Catalog the Data

DataCatalog

DataCatalog

LegacyMetadata

LegacyMetadata

ExternalMetadata(B2B)

ExternalMetadata(B2B)

Build Information Model

InformationModel

InformationModel

Reference Architecture

Data

Data Services/Abstraction

Create a Service Model

CandidateServices

CandidateServicesUnderstand Services

Information to Services

ServicesAndInformation

ServicesAndInformation

Build Service Model

ServiceModel

ServiceModel

DataCatalog

DataCatalog

InformationModel

InformationModel

Reference Architecture

Data

Data Services/Abstraction

Transactional Services

Cloud Provider

On Premises

Cloud Provider

On Premises

Create a Process Model

CandidateProcesses

CandidateProcessesUnderstand Processes

Services to Processes

Services toProcesses

Services toProcesses

Build Process Model

ProcessModel

ProcessModel

DataCatalog

DataCatalog

InformationModel

InformationModel

ServiceModel

ServiceModel

Reference Architecture

Data

Data Services/Abstraction

Transactional Services

Process Management (BPMS) Composites/PortalsRules Management

On Premises Cloud Provider

Create a Governance Model

Defined Policies

Defined PoliciesDefine Policies

Design Policies

PolicyDesigns

PolicyDesigns

Implement Policies

GovernanceModel

GovernanceModel

ProcessModel

ProcessModel

InformationModel

InformationModel

ServiceModel

ServiceModel

Reference Architecture

Security & Identity

Managem

ent & Service G

overnance

Data

Data Services/Abstraction

Transactional Services

Process Management (BPMS) Composites/Portals

Monitoring and Management

Rules Management

RuntimeGovernance

Repository Logs

Policies

Policies

Monitoring

Test SOA using CloudArchitecture

TestPlan

TestPlanCreate Test Plan

Black Box Testing

Black BoxTestingResults

Black BoxTestingResults

White Box Testing

White BoxTestingResults

White BoxTestingResults

ProcessModel

ProcessModel

InformationModel

InformationModel

ServiceModel

ServiceModel

GovernanceModel

GovernanceModel

Assigning Candidate Data, Services, and Processes for the Clouds

ProcessAssignments

ProcessAssignments

Analyze Candidate Processes

Analyze Candidate Services

ServiceAssignments

ServiceAssignments

Analyze Candidate Data

DataAssignments

DataAssignments

CandidateProcesses

CandidateProcesses

CandidateServices

CandidateServices

Data Dictionaryand Metadata

Data Dictionaryand Metadata

GovernanceModel

GovernanceModel

Select Platforms and DeployProcesses, Services, and Datato Platforms.

CandidatePlatforms

CandidatePlatforms

List Candidate Platforms

Analyze and Test Candidate Platforms

TestResults

TestResults

Select Target Platforms

TargetPlatforms

TargetPlatforms

ProcessAssignments

ProcessAssignments

ServiceAssignments

ServiceAssignments

DataAssignments

DataAssignments

Deploy toTarget Platforms

Data

Data

Data

Data

Data Services Processes

Data

Data

Data

Data

Data Services Processes

PaaS

IaaSDaaS

Data

Data

Data

Data

Data Services Processes

PaaS

IaaS

DaaS

Application

Application

Stepping to the Clouds

1. Access the mission.2. Access the culture.3. Access the value.4. Understand your data.5. Understand your services.6. Understand your processes.7. Understand the cloud

resources.8. Identify candidate data.9. Identify candidate services10. Identify candidate processes.

11. Create a governance strategy.12. Create a security strategy. 13. Bind candidate services to data

and processes.14. Relocate services, processes,

and information. 15. Implement security. 16. Implement governance.17. Implement operations.

Q&A

We are leading the conversation with thought leadership and cutting edge services.

The shift to Cloud

#1 BEST SELLING BOOK ON CLOUD COMPUTING

#1 MOST READ BLOG ON CLOUD COMPUTING

#1 MOST LISTENED TO PODCAST ON CLOUD COMPUTING WITH OVER 10,000 LISTENERS

Additional Slides

Use Case: Creating an Hybrid Cloud

PublicCloud

PrivateCloud

PublicCloud

PrivateCloud

API/Services

Migration

Interoperability

Current Patterns: Primitive Hybrid Architectures

Best: Cloud Abstraction

PrivateCloud B

PublicCloud C

PrivateCloud A

Abstraction Layer(Composite)

Emerging Hybrid Cloud Architectural Patterns

• Static Placement• Assisted

Replication• Auto Migration• Dynamic Migration

82

Static Placement

83

Public Cloud

Data Data

Private Cloud

Internet

Data Data

Services/APIBind toPlatform

- Difficult or impossible to dynamically port or migrate. - Little or no use of standards. - Typically older platforms mixed with cloud platforms. - Little use of abstraction.

Assisted Replication

84

Public Cloud

Data Data

Private Cloud

Internet

Data Data

Services/API

- Code port and/or interface compatibility. - Some use of standards. - Typically new platforms that are code/service compatible.- Limited use of abstraction.

Auto Migration

85

Public Cloud

Data Data

Private Cloud

Internet

Data Data

Services/API

- Auto code/VM migration and/or interface compatibility. - Some use of standards. - Virtualization, code, and data levels.- Some use of abstraction.

Dynamic Migration

8686

Public Cloud

Data Data

Private Cloud

Internet

Data Data

Services/API

- Dynamic VM migration within single logical OS.- Logical and physical abstraction. - Some use of standards. - Hybrid cloud operating systems.

OS

RuntimeGovernance

Repository Logs

Policies

Policies

Monitoring

We Need Cloud Abstraction!

PrivateCloud B

PublicCloud C

PrivateCloud A

CommonInterface

Cloud Management

Define Information

89

Data DataCore Physical and Logical Models

90

Define Services

91

Services

Data DataCore Physical and Logical Models

Core Services

Define Processes

93

Services

Data DataCore Physical and Logical Models

Core Services

Core Processes

Map to the Proper Architecture

• Static Placement• Assisted

Replication• Auto Migration• Dynamic Migration

94

Services

Data Data

Select Technology Solution

• Static Placement• Assisted

Replication• Auto Migration• Dynamic Migration

95