Modern Computer Systems, Clusters, And Networks

Chapter 11 Modern Computer Systems, Clusters, and Networks

Basic Personal Computer System

Chapter 11: Modern Computer Systems, Clusters, and Networks

11-2

Mainframe Computer System

Chapter 11: Modern Computer Systems, Clusters, and Networks

11-3

Major PC System Components

Chapter 11: Modern Computer Systems, Clusters, and Networks

11-4

System Performance Improvements     

Multiple CPUs Faster clock speed, buses and circuits Wider instruction and data paths Faster disk access More and faster memory

Chapter 11: Modern Computer Systems, Clusters, and Networks

11-5

Multiprocessing  Reasons  Increase the processing power of a system  Parallel processing

 Types of multiprocessor systems  Tightly coupled systems  Loosely coupled systems

Chapter 11: Modern Computer Systems, Clusters, and Networks

11-6

Tightly Coupled Systems  Also called multiprocessor systems  Identical access to programs, data, shared memory, I/O, etc.  Easily extends multi-tasking, and redundant program execution  Two ways to configure  Master-slave multiprocessing  Symmetrical multiprocessing (SMP) Chapter 11: Modern Computer Systems, Clusters, and Networks

11-7

Tightly Coupled Systems

Chapter 11: Modern Computer Systems, Clusters, and Networks

11-8

Master-Slave Multiprocessing  Master CPU  Manages the system  Controls all resources and scheduling  Assigns tasks to slave CPUs

 Advantages  Simplicity  Protection of system and data

 Disadvantages  Master CPU becomes a bottleneck  Reliability issues – if master CPU fails entire system fails Chapter 11: Modern Computer Systems, Clusters, and Networks

11-9

Symmetrical Multiprocessing  Each CPU has equal access to resources  Each CPU determines what to run using a standard algorithm  Disadvantages  Resource conflicts – memory, i/o, etc.  Complex implementation

 Advantages  High reliability  Fault tolerant support is straightforward  Balanced workload Chapter 11: Modern Computer Systems, Clusters, and Networks

11-10

Loosely Coupled Systems  Clusters or multi-computer systems  Each system has its own CPU, memory, and I/O facilities  Each system is known as a node of the cluster  Advantages  Fault-tolerant, scalable, well balanced, distance is not an issue

 Two ways to configure  Shared-nothing model  Shared-disk model Chapter 11: Modern Computer Systems, Clusters, and Networks

11-11

Shared-Nothing Model  High speed link between nodes  No sharing of resources  Partitioning of work through division of data  Advantage  Reduced communication between nodes

 Disadvantage  Can result in inefficient division of work Chapter 11: Modern Computer Systems, Clusters, and Networks

11-12

Shared-Disk Model  High speed link between nodes  Disk drives are shared between nodes  Advantage  Better load balancing

 Disadvantage  Complex software required for transactional processing (lock, commit phases) Chapter 11: Modern Computer Systems, Clusters, and Networks

11-13

Cluster Models

Chapter 11: Modern Computer Systems, Clusters, and Networks

11-14

Beowulf Clusters  Simple and highly configurable  Low cost  Networked  Computers connected to one another by a private Ethernet network  Connection to an external network is through a single gateway computer

 Configuration  COTS – Commodity-off-the-shelf components such as inexpensive computers  Blade components – computers mounted on a motherboard that are plugged into connectors on a rack  Either shared-disk or shared-nothing model Chapter 11: Modern Computer Systems, Clusters, and Networks

11-15

Blade and Rack of Beowulf Cluster

Figure 11.9 Chapter 11: Modern Computer Systems, Clusters, and Networks

11-16

Computer Interconnection  Communication channel – pathway for data movement between computers  Point-to-Point connectivity  Communication channel that passes data directly between two computers  Serial connection  Telephone modem  Terminal controller – handles multiple point-to-point connections for a host computer

 Multipoint connectivity  Multidrop channel or shared communication channel Chapter 11: Modern Computer Systems, Clusters, and Networks

11-17

Example: Point-to-Point

Chapter 11: Modern Computer Systems, Clusters, and Networks

11-18

Client-Server Architecture  Computer servers provides services  File storage, databases, printing services, login services, web services

 Client computers  Execute programs in its own memory  Access files either locally or can request files from a server

Chapter 11: Modern Computer Systems, Clusters, and Networks

11-19

Client-Server Network

Chapter 11: Modern Computer Systems, Clusters, and Networks

11-20

LAN Topology  Arrangement of workstations in a shared medium environment  Logical arrangement (data flow)  Physical arrangement (cabling scheme)

Chapter 11: Modern Computer Systems, Clusters, and Networks

11-21

LAN Topologies: Bus  Multipoint medium  Stations attach to linear medium (bus) using tap  Transmission from any stations travels entire medium (both directions)  Termination required at ends of bus to prevent the signal from bouncing  Break in cable brings down entire bus Chapter 11: Modern Computer Systems, Clusters, and Networks

11-22

Bus LAN Diagram

Chapter 11: Modern Computer Systems, Clusters, and Networks

11-23

LAN Topologies: Tree  Generalization of bus topology  Branching cable with no closed loops  Cable(s) begin at headend, travel to branches which may have branches of their own  Each transmission propagates through network, can be received by any station Chapter 11: Modern Computer Systems, Clusters, and Networks

11-24

LAN Topologies: Ring  Repeaters are joined by unidirectional point-to-point links in a ring  As data circulates past a receiver, the receiver checks its address, and copies those intended for it into a local buffer  Data circulates until it returns to source, which removes it from network  Better performance at high levels of usage Chapter 11: Modern Computer Systems, Clusters, and Networks

11-25

Ring LAN Diagram

Chapter 11: Modern Computer Systems, Clusters, and Networks

11-26

LAN Topologies: Star  Each station connected point-to-point to a central station, usually with two undirectional links  Switching in the central station connects pairs of nodes together  Central node can broadcast info, or can switch frames among stations  Failure of central station causes entire network to go down Chapter 11: Modern Computer Systems, Clusters, and Networks

11-27

Star LAN Diagram

Chapter 11: Modern Computer Systems, Clusters, and Networks

11-28

Ethernet MAC Protocol  MAC – Medium Access Control  Ethernet and CSMA/CD  Carrier sense multiple access with collision detection

 Four step procedure    

If medium is idle, transmit If medium is busy, listen until idle and then transmit If collision is detected, cease transmitting After a collision, wait a random amount of time before retransmitting

Chapter 11: Modern Computer Systems, Clusters, and Networks

11-29

Ethernet Frame

Chapter 11: Modern Computer Systems, Clusters, and Networks

11-30

Switched Ethernet

Chapter 11: Modern Computer Systems, Clusters, and Networks

11-31

Token Ring MAC Protocol  Token “seized” by changing a bit on the circulating frame to indicate start of frame rather than token  Default configuration requires sender to complete transmission and begin receiving transmitted frame before releasing the token  “Early token release” allows release of token after transmission but before receipt of frame Chapter 11: Modern Computer Systems, Clusters, and Networks

11-32

Hubs  The active central element of the star layout.  When a single station transmits, the hub repeats the signal on the outgoing line to each station.  Hubs can be cascaded in a hierarchical configuration  Ethernet hubs are physically a star but logically a bus. Chapter 11: Modern Computer Systems, Clusters, and Networks

11-33

Bridges  Allow connections between LANs and to WANs  Used between similar networks  Read all frames from each network  Accept frames from sender on one network that are addressed to a receiver on the other network  Retransmit frames from sender using MAC protocol for receiver Chapter 11: Modern Computer Systems, Clusters, and Networks

11-34

Routers  Similar to bridges but connect dissimilar networks  Convert format of the message to correspond to the protocol of the other network  Network traffic is specifically addressed to the router

Chapter 11: Modern Computer Systems, Clusters, and Networks

11-35

Wide Area Network  Circuit switching  Dedicated channel between source and destination for duration of connection

 Message switching  Dedicated channel for an entire message

 Packet switching  An independent path is created for each datagram

 Virtual circuit switching  A route is created from source to destination before transmission begins and all datagrams are sent using the same route Chapter 11: Modern Computer Systems, Clusters, and Networks

11-36

Networks vs. Clusters  Externally, clusters appear as a single computing unit.  Network nodes are individually identifiable.  Workload on a cluster is determined by cluster administration and loadbalancing software.  Network workload cannot be controlled using the above method. Chapter 11: Modern Computer Systems, Clusters, and Networks

11-37

High Performance Computing  Massively parallel processor architectures (MPP)  Clusters of power machines or larger Beowulf blade clusters  Well suited for problems that can be broken into subtasks

 Grid computing  Supercomputer performance through distributing CPU processing to the spare CPU cycles of personal computers connected to a network Chapter 11: Modern Computer Systems, Clusters, and Networks

11-38

Parallel Computers    

Massively parallel architectures Hundreds to millions of CPUs CPUs have small amounts of local memory All CPUs have access to global shared memory  Pipelined CPUs  Results from one CPU flow to the next CPU for additional processing

Chapter 11: Modern Computer Systems, Clusters, and Networks

11-39