Computação nas Nuvens (Cloud Computing)

Departamento de Informática e Estatística
Centro Tecnológico (CTC)

Andando nas Nuvens

Prof. Mario Dantas, PhD
LaPeSD
(Laboratório de Pesquisas em Sistemas Distribuídos)
mario@inf.ufsc.br

“Quando a rede externa é tão rápida quanto aos enlaces internos
dos computadores, as configurações dos computadores são
desintegradas ao longo da rede na forma de um conjunto de
dispositivos especiais específicos”

(George Gilder, 2001)

Arquitetura de Computadores
Convencionais

Arquitetura de von Neumann

CPU
memória
Barramento
Principal

Barramento
saída entrada de E/S

Arquitetura dos Sistemas Modernos

Remote Viz.
Groupware
Virtualization
Web Portal

Collaboration

PC, Laptop, Graphic Wkstn
PDA, .... etc. Render Engine

Lab or Field Interpretation
Data collection - visualization

[Dominic Lam, IBM]

Responsabilidades Básicas

Alarme
Circuito Interno
de incêndio
TV
Armário
Cabeamento
Vertical

Fax
Roteador
Wireless
Computadores

3° Piso
Armário
Cabeamento
Vertical
Impressora
Câmera uso geral
Patch vigilância
Backbone

Panel 2° Piso
Armário da
Rede Controle

Armário da Rede
Armário da Computadores
Rede
Telefônica
Painel de Conjunto
Servidores 1° Piso
Controle
de Dados

PABX

[Dantas, 2009]

Cabo coaxial da
spliter
operadora TV Cabo

Cable
box Cable modem

Junction
box

Antena

Head
End

Gateway
Internet

Servidores de [Dantas, 2009]
Serviços Internet


[Dantas, 2009]
[Furukawa, 2009]


Estação de
Trabalho
Servidores de
Banco de Dados

Computador de
Mainframe uso especial

[Dantas, 2009]


[Dantas, 2009]

Responsabilidades Modernas

Computação nas Nuvens

Agenda

 Motivação

 Conceitos Internet, Web, Cluster e Grid Computacionais

 Computação nas Nuvens

 Estudos de Casos

 Conclusões

Motivação

 Computação nas Nuvens: O futuro, segundo o Google!
[http://www.undergoogle.com/blog]

Microsoft investe em computação em nuvem
[InformationWeek EUA,10/11/2008]

 Dell oference soluções diferenciadas para as nuvens
[http://www.dell.com/cloudcomputing]

 Amazon oferece serviço de nuvem
[http://aws.amazon.com/ec2/]

Motivação
 Vender mais aplicativos

 Criar aplicativos sensacionais com maior rapidez

 Fornecer qualquer aplicativo de negócios

 Alcançar a lucratividade mais rapidamente

Credibilidade instantânea

 Tirar proveito da nossa rede

[http://www.salesforce.com/br/platform/partners.jsp]

Motivação
Noticia de revista semanal: 12 de Agosto de 2009

Conceitos

 Internet
 Grid
 Web
 Cloud
 Cluster

Qual a diferença entre:

Conceitos
> História da Internet Setembro de 1973

Conceitos

> História da WEB

Where the web was born
Tim Berners-Lee, a scientist at CERN, invented the World Wide Web (WWW) in 1989.
The Web was originally conceived and developed to meet the demand for automatic
information sharing between scientists working in different universities and institutes
all over the world.

Conceitos
> História dos Clusters

Conceitos


[Dantas, 2005]

Conceitos


Conceitos

Rede de
Cray X-MP Comunicação

Tesla Nvidia (GPU)

Laptop
Playstation 3
Cell Processor

Conceitos

Architecture Count Share % Processor Sum
Constellations 2 0.40 % 17.648
MPP 88 17.60 % 1.964.865
Cluster 410 82.00 % 2.122.220
Totals 500 100%

Participação das Arquiteturas em 06/2009
no TOP500 [http://www.top500.org]

Conceitos

> História dos Grids

Ciência Empresas Indivíduos
Conectividade

Grid Cloud ???

Tempo

(Ian Foster, WorldComp, 2009)


O conceito de cloud incorpora os componentes:

• Infra-estrutura (Infrastructure as a service – IaaS);

• Plataformas (Platform as a service – PaaS); e

• Pacotes de software (Software as a Service - SaaS)
como serviços.

Várias empresas já oferecem ferramentas para a implementação
de cloud computing, exemplos são:

• Google,
• Microsoft,
• Salesforce;
• Yahoo

Por outro lado, algumas empresas já oferecem serviços como
ambientes de cloud, exemplos são:

• Amazon;
• GE;
• L’Óréal.

Microsoft
AMAZON

ESTUDOS DE CASOS HP

DE
NUVENS COMPUTACIONAIS
Dell Yahoo

Google

(A) Yahoo Cloud Computing

Cloud computing is becoming an important arena for researchers
and developers to test next generation software services, and
scalable software and systems is at the heart of providing cloud
services via the Internet.

[http://labs.yahoo.com/Cloud_Computing]

(A) Yahoo Cloud Computing

At Yahoo!, we are very supportive of academic research in cloud
computing that, to date, has been limited due to significant cost
barriers in getting large computing systems operational and the
lack of software tools.

By making large systems available and by contributing to open
source software such as Pig and Hadoop, we hope to enable
researchers and students to innovate and create new kinds of
systems, applications, and tools, pushing the boundaries in this
field.

[http://labs.yahoo.com/Cloud_Computing]

(B) Microsoft Cloud Computing Infrastructure

The flexibility and agility you need to deliver an on-premise
'private' cloud for your infrastructure, applications, management
and security.

Cloud Computing Infrastructure

What Makes Up a Cloud

The principles of cloud computing include:

• Virtualization and automation;
• Interchangeable (fungible) resources such as servers;
• Storage; and
• Network

(B) Microsoft Cloud Computing Infrastructure

Management of these resources as a single fabric Elastic
capacity (scale up or down) to respond to business demands
Applications (and the tools to develop them) that can
truly scale out.

Focused on the service delivered to the business

(C) LaPeSD 1

A new proposed architecture
[Dantas et al., 2009]

Experimental Environment and Results
- The ontology components were built utilizing Protégé 3.3.1 and the multi-cluster environment
was simulated using the SimGrid software package;

-The goal of the SimGrid project is to facilitate research in the area of distributed and parallel
application scheduling on distributed computing platforms, ranging from a simple network of
workstations to large computational grids;

-The SimGrid toolkit provides core functionalities for the simulation of distributed applications
in heterogeneous distributed environments;

-This characteristic of the package is interesting since it enables simulation of multi-cluster
configurations;
VO-A # processors VO-B # processors

Cluster_01 7 Cluster_1 4

Cluster_02 5 Cluster_2 4

Cluster_03 3 - -

Multi-cluster configuration



- After creating the multi-cluster configuration, we identified a number of basic concepts
to be used in both virtual organizations for requesting resources;

- Table 2 presents how these request concepts are understood in virtual organizations A and B.
Virtual organization A represents a foreign environment, whereas virtual B is a Brazilian
configuration.

The ontology set utilized for the experiments


- The first test was a query where a user required a certain minimum amount physical
memory, operating system type (UNIX) and number of processors equal to 2.

The first test without dynamic information


- Figure 6 shows the second test, where the same query was executed under the new
system;
- In addition to the available environment, this figure presents numbers related to the
number of processors in each node and link workload to server.

The first query in the new environment



-The third experiment represents a query as it was in the original environment, where a user requires 6
processors. Figure 7 shows the result for this query;
- As expected, the system replies that no match was found. This answer is based upon the static
information available in the original system.

The third experiment without a dynamic information


- The next experiment represents the previous one considering all new features of the new proposal;
- A comprehensive picture of available multi-clusters is presented in Figure 8;
- This figure shows the same third query, however now executed under the new proposal.

The third experiment under the new environment

(D) LaPeSD 2: Mobile Access and Interaction
(2b) – Submeter (2a) – Coordenar
tarefas tarefas
Tarefa 1
Tarefa 2
Escalonador Gerenciador
Tarefa 3
(Condor) Workflow
Tarefa 4

(1)Submissão
Agente (4) Está conectado?
(5) Resposta


(D) LaPeSD 2: Mobile Access and Interaction


(E) Amazon Elastic Compute Cloud (Amazon EC2)

Amazon Elastic Compute Cloud (Amazon EC2) is a web service that provides resizable
compute capacity in the cloud. It is designed to make web-scale computing easier for
developers.

Amazon EC2’s simple web service interface allows you to obtain and configure capacity
with minimal friction.

It provides you with complete control of your computing resources and lets you run
on Amazon’s proven computing environment.


(E) Amazon Elastic Compute Cloud (Amazon EC2)

Amazon EC2 reduces the time required to obtain and boot new server instances to
minutes, allowing you to quickly scale capacity, both up and down, as your computing
requirements change. Amazon EC2 changes the economics of computing by
allowing you to pay only for capacity that you actually use. Amazon EC2 provides
developers the tools to build failure resilient applications and isolate themselves
from common failure scenarios.


Visão de Computação nas Nuvens:

- Abordagem de sistemas distribuídos;

- Fortemente baseada em redes geograficamente
distribuídas;

- Modelo semelhante as grades computacionais para
ambientes comerciais;
- Novo paradigma comercial de utilização de recursos e
serviços na web.

Referências Bibliográficas

[Dantas, 2002] [Dantas, 2005] [Dantas, 2009]

Computação nas Nuvens (Cloud Computing)

Recommended

Recommended

More Related Content

Featured

Featured (20)

Computação nas Nuvens (Cloud Computing)