Transcript Computer Systems and Networks

ITEC 1000 “Introduction to Information Technology”
Lecture 11:
Computer Systems and Networks
1
Lecture Template:
System Performance Improvement
 Multiprocessing
 Computer Interconnection
 OSI model and TCP/IP
 Network Topologies
 Protocols
 Wide Area Network
 High Performance Computing

2
System Performance
Improvements

Multiple CPUs
Sharing memory and I/O facilities
Conflict among the CPUs for shared resources

Faster clock speed, buses and circuits
Improving technology to design faster CPU circuits and
buses

Wider instruction and data paths
Wider interface between the CPU and memory allows to
fetch more data in a single operation

Faster disk access
Smaller discs, more density packed, increased storage
RAID: data on different discs, multiple access
simultaneously

More and faster memory
Reduces the time to access instructions and data
3
Multiprocessing

Reasons
Increase the processing power of a
system
Parallel processing

Types of multiprocessor systems
Tightly coupled systems
Loosely coupled systems
4
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)
5
Tightly Coupled Systems
6
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
7
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
8
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
9
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
10
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)
11
Cluster Models
12
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-topoint connections for a host computer

Multipoint connectivity
Multidrop channel or shared communication
channel
13
Example: Point-to-Point
14
Definitions
Topology: the way in which loosely coupled
computers are interconnected
Synonym: configuration
Protocol: a set of rules and standards for
communications between computers
15
Client-Server Architecture

Computer server 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
16
Client-Server Network
F
server
17
A typical data packet
18
The layers of the OSI model
19
Passing a message through an
intermediate node
20
A comparison of OSI and TCP/IP
21
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
22
Blade and Rack of Beowulf Cluster
23
LAN Topology
Arrangement of workstations in a
shared medium environment
 Logical arrangement (data flow)
 Physical arrangement (cabling
scheme)

24
Network Topologies
Bus:
Nodes are connected to a common bus with a
terminator on each end.
Ring:
Similar to bus, but bus is closed - no ends.
Star:
A central node does most of the processing. Remote
nodes are connected point-to-point with it.
Loop:
Similar to ring but nodes are directly in
communication path.
Hierarchical: A tree-like structure emanating from a central or
root node.
Web:
Everything connected to everything else.
25
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
26
Ring LAN Diagram
27
Ring Topology
28
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

29
Bus LAN Diagram
30
Bus Topology
31
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

32
LAN Topologies: Star




Each station connected point-to-point to a
central station, usually with two
unidirectional 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
33
Star LAN Diagram
34
Star Topology
Guy in
the
Middle
35
Loop Topology
36
Hierarchical Topology
Big Guy at Top
37
Web Topology
38
Topology Selection: Factors
Distances between stations
 Layout of the room/building
 Overall size of the network
 Distance between the most remote
nodes
 Speed requirements
 Network traffic
 Total number of stations

39
Medium Access Control (MAC)
Protocols
Characteristics of the channels, data
rate, voltage levels, etc.
 Node access to the channel (medium
access control protocol)

Steer data to its destination
Detect errors
Prevent multiple nodes from accessing
the network simultaneously (collision)
Ethernet and token ring
Implemented in hardware
40
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
41
Ethernet Frame
42
Switched Ethernet
43
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
44
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

45
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
46
Gateways
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

47
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
48
A Switched Wide Area Network
49
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 50

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
51
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
52