Transcript PowerPoint 프레젠테이션

1.3 System Models for Distributed and Cloud Computing
Classification of Massive systems (Table 1.2)
1.3.1 Clusters of Cooperative Computers
 Cluster Architecture
 The architecture of a typical server cluster built around a low-latency, highbandwidth interconnection network. (Fig. 1.15)
 Single-System Image
 An ideal cluster should merge multiple system images into a single-system
image (SSI).
 cluster operating system
 some middleware to support SSI at various levels.
 An SSI is an illusion created by software or hardware that presents a collection of
resources as one integrated, powerful resource.
 Critical Cluster Design Issues and Feasible Implementations (Table 1.3)
1.3.2 Grid Computing Infrastructure
 Computational Grids
 Computational grid and data grid providing computing utility, data, and
information services through resource sharing and cooperation among
participating organizations. (Fig. 1.16)
Ch. 1-2
1
 Grid Families
 Grid systems are classified in essentially two categories: computational or data
grids and P2P grids. (Table 1.4)
1.3.3 Peer-to-Peer Network Families
 P2P Systems
 In a P2P system, every node acts as both a client and a server, providing part
of the system resources.
 The physical network is simply an ad hoc network formed at various Internet
domains randomly using the TCP/IP and NAI (Network Access Identifier)
protocols.
Ch. 1-2
2
 Overlay Networks
 The overlay is a virtual network formed by mapping each physical
machine with its ID, logically, through a virtual mapping as shown in Fig.
1.17.
 There are two types of overlay networks: unstructured and structured.
 P2P Application Families
 Major Categories of P2P Network Families. (Table 1.5)
 P2P Computing Challenges
 P2P computing faces three types of heterogeneity problem in hardware,
software, and network requirements.
 P2P performance is affected by routing efficiency and self-organization
by participating peers.
 Fault tolerance, failure management, and load balancing are other
important issues in using overlay networks.
 Security, privacy, and copyright violations are major worries by those in
the industry in terms of applying P2P technology in business application.
Ch. 1-2
3
1.3.4 Cloud Computing Over the Internet
 Internet Clouds
 Cloud computing applies a virtualized platform with elastic resources on
demand by provisioning hardware, software, and data set dynamically. (Fig.
1. 18)
 The Cloud Landscape
 Three cloud service models in a cloud landscape of major providers. (Fig.
1.19)
Ch. 1-2
4
1.4 Software Environments for Distributed Systems and Clouds
1.4.1 Service-Oriented Architecture (SOA)
 대규모 컴퓨터 시스템을 구축할 때의 개념으로 업무상에 일 처리에 해당하는
소프트웨어 기능을 서비스로 판단하여 그 서비스를 네트워크상에 연동하여
시스템 전체를 구축해 나가는 방법론.
Ch. 1-2
5
 The Evolution of SOA
 SOA applies to building grids, clouds, grid of clouds, clouds of grids, clouds
of clouds, and systems of system in general.
 The evolution of SOA: grids of clouds and grids. (Fig. 1.21)
 SOA aims to search for, or sort out, the useful data from the massive amount
of raw data items.
 Grids versus Clouds
 In general, a grid system applies static resources, while a cloud emphasizes
elastic resources.
 The difference between grids and clouds are limited only in dynamic resource
allocation based on virtualization and autonomic computing.
 Trends toward Distributed Operating Systems
 DOS achieves higher use or system transparency.
 A transparent computing environment that separates the user data, OS, and
hardware in time and space – an ideal model for cloud computing. (Fig. 1.22)
Ch. 1-2
6
1.5 Performance, Security and Energy Efficient
 Performance Metrics and Scalability Analysis
 Performance Metrics
 MIPS
 Mbps
 Tflops (tera floating-point operations per second)
 TPS (transactions per second)
 job response time
 network latency
 Dimensions of Scalability
 Size scalability
 Software scalability
 Application scalability
 Technology scalability
Ch. 1-2
7
 Fault Tolerance and System Availability
 System Availability
 A system is highly available if it has long mean time to failure (MTTF) and a short
mean time to repair (MTTR).
 System Availability=MTTF/(MTTF+MTTR)
 MTTF는 주어진 시간에서 고장 발생시 까지 시간으로 고장 수리 후 다음 고장까지의
시간을 의미함
 Any failure that will be pull down the operation of the entire system is called a
single of failure. The rule of thumb is to design a dependable computting system
with no single point of failure.
 Network Threats and Data Integrity
 Threats to Systems and Networks
 Fig. 1.25 summaries various attack types and their potential damages
to users.
 Security Responsibilities
 Three security requirements are often considered: confidentiality,
integrity, and availability for most Internet service providers and cloud
users.
Ch. 1-2
8
 Energy Efficiency in Distributed Computing
 Parallel and distributed computing systems recently encountered
new challenging issues including energy efficiency, and workload
and resource outsourcing.
 클라우드 컴퓨팅은 IT자원을 외부에 아웃소싱을 함으로 인하여 가장 먼저 대두 되
는 것이 ‘보안’에 관련된 문제이다.
Ch. 1-2
9