Download DistSys and more Slides Computer Networks in PDF only on Docsity!
1
Introduction to Distributed Systems CIS 505: Software Systems
Insup Lee
Department of Computer and Information Science
University of Pennsylvania
CIS 505, Spring 2007
CIS 505, Spring 2007
Distributed Systems
2
Distributed Systems
Why distributed systems?
o availability of powerful yet cheap microprocessors (PCs, workstations, PDAs, embedded systems, etc.)
o continuing advances in communication technology
What is a distributed system?
o A distributed system is a collection of independent coherent system.computers that appear to the users of the system as a single
CIS 505, Spring 2007
Distributed Systems
3
Examples
sharingThe world wide web โ information, resource
Clusters, Network of workstations
automated assembly line)Distributed manufacturing system (e.g.,
processing of ordersInformation system to handle automaticNetwork of branch office computers -
Network of embedded systems
New Cell processor (PlayStation 3)
CIS 505, Spring 2007
Distributed Systems
4
Advantages and disadvantages
Advantages
o Incremental growtho Reliabilityo Inherent distributiono Speedo Economics
Disadvantages
o Securityo More components to failo Networko Software
CIS 505, Spring 2007
Distributed Systems
5
Organization of a Distributed System
Note that the middleware layer extends over multiple machines. A distributed system organized as middleware.
CIS 505, Spring 2007
Distributed Systems
6
Goals of Distributed Systems
Transparency
Openness
Reliability
Performance
Scalability
CIS 505, Spring 2007
Distributed Systems
7
Transparency
appear as a single computer.how to make a collection of computersHow to achieve the single-system image, i.e.,
2) at a lower level, make the system look1) hide the distribution from usersachieved at two levels:well as the application programs can beHiding all the distribution from the users as
transparent to programs.
1) and 2) requires uniform interfaces such as access
to files, communication.
CIS 505, Spring 2007
Distributed Systems
8
Transparency in a Distributed System
Different forms of transparency in a distributed system.
Hide whether a (software) resource is in memory or on disk
Persistence
Hide the failure and recovery of a resource
Failure
usersHide that a resource may be shared by several competitive
Concurrency
usersHide that a resource may be shared by several competitive
Replication
in useHide that a resource may be moved to another location while
Relocation
Hide that a resource may move to another location
Migration
Hide where a resource is located
Location
accessedHide differences in data representation and how a resource is
Access
Description
Transparency
CIS 505, Spring 2007
Distributed Systems
13
Scalability Problems
Characteristics of decentralized algorithms:
the system state.No machine has complete information about
local information.Machines make decisions based only on
algorithm.Failure of one machine does not ruin the
clock exists.There is no implicit assumption that a global
CIS 505, Spring 2007
Distributed Systems
14
Scaling Techniques (1)
a) The difference between letting:
a server or
b)
a client check forms as they are being filled
CIS 505, Spring 2007
Distributed Systems
15
Scaling Techniques (2)
An example of dividing the DNS name space into zones.
CIS 505, Spring 2007
Distributed Systems
16
Pitfalls when Developing Distributed Systems
False assumptions made by first time developer:
The network is reliable.
The network is secure.
The network is homogeneous.
The topology does not change.
Latency is zero.
Bandwidth is infinite.
Transport cost is zero.
There is one administrator.
CIS 505, Spring 2007
Distributed Systems
17
Hardware Concepts
computer systems: multiprocessors vs. multicomputersDifferent basic organizations and memories in distributed
CIS 505, Spring 2007
Distributed Systems
18
Multiprocessors (1)
A bus-based multiprocessor
o Cache memory, hit rate, coherence, write-through
cache, snoopy cache
CIS 505, Spring 2007
Distributed Systems
19
Multiprocessors (2)
a)
A crossbar switch
b)
An omega switching network
CIS 505, Spring 2007
Distributed Systems
20
Homogeneous Multicomputer Systems
a)
Grid
b)
Hypercube
Tightly coupled vs. loosely coupled
CIS 505, Spring 2007
Distributed Systems
25
Types of Distributed Systems
Distributed Computing Systems
Distributed information systems
Distributed Pervasive/Embedded Systems
CIS 505, Spring 2007
Distributed Systems
26
Cluster Computing Systems
Figure 1-6. An example of a cluster computing system.
CIS 505, Spring 2007
Distributed Systems
27
Grid Computing Systems
systems.Figure 1-7. A layered architecture for grid computing
CIS 505, Spring 2007
Distributed Systems
28
Transaction Processing Systems (1)
Figure 1-8. Example primitives for transactions.
CIS 505, Spring 2007
Distributed Systems
29
Transaction Processing Systems (2)
Characteristic properties of transactions:
transaction happens indivisibly.Atomic: To the outside world, the
violate system invariants.Consistent: The transaction does not
interfere with each other.Isolated: Concurrent transactions do not
the changes are permanent.Durable: Once a transaction commits,
CIS 505, Spring 2007
Distributed Systems
30
Transaction Processing Systems (3)
Figure 1-9. A nested transaction.
CIS 505, Spring 2007
Distributed Systems
31
Enterprise Application Integration
enterprise application integration.Figure 1-11. Middleware as a communication facilitator in
CIS 505, Spring 2007
Distributed Systems
32
Distributed Pervasive Systems
Requirements for pervasive systems
Embrace contextual changes.
Encourage ad hoc composition.
Recognize sharing as the default.
CIS 505, Spring 2007
Distributed Systems
37
Desired Trends
Volume &
Diversity Quality &
Unit cost Usability
Maintenance
cost
CIS 505, Spring 2007
Distributed Systems
38
Electronic Health Care Systems (1)
Questions to be addressed for health care systems:
stored?Where and how should monitored data be
How can we prevent loss of crucial data?
propagate alerts?What infrastructure is needed to generate and
How can physicians provide online feedback?
system be realized?How can extreme robustness of the monitoring
proper policies be enforced?What are the security issues and how can the
CIS 505, Spring 2007
Distributed Systems
39
Electronic Health Care Systems (2)
health care system, using (a) a local hub orFigure 1-12. Monitoring a person in a pervasive electronic
(b) a continuous wireless connection.
CIS 505, Spring 2007
Distributed Systems
40
Background: Sensor Networks
Types of sensors
o Seismic, low sampling rate magnetic, thermal, visual, infrared, acoustic and radar
Conditions to monitor
o Temperature, humidity, (vehicular) movement, lightning condition, pressure, soil makeup, noise levels
o Current characteristics such as speed, direction, and size of ano Mechanical stress levels on attached objectso Presence or absence of certain kinds of objects object
CIS 505, Spring 2007
Distributed Systems
41
Technology Trends: Mote Evolution
CIS 505, Spring 2007
Distributed Systems
42
SN Characteristics
Environment
o connect to physical environment (large numbers, dense, real-time)
o Sensor nodes are prone to failures, non-deterministic
o wireless communication
o massively parallel interfaces (to users and applications)
o management critical)Limited resources: battery, bandwidth, memory, CPU (power
Network
o Topology changes dynamically
o sporadic connectivity
o new resources entering/leaving
o large amounts of redundancy
o self-configure/re-configure
CIS 505, Spring 2007
Distributed Systems
43
SN applications
Infrastructure security, military applications
Environmental and Habitat monitoring
Health applications
Smart space/home applications
Other commercial applications
o Interactive museumso Traffic Control, vehicle tracking and detectiono Industrial Sensing
CIS 505, Spring 2007
Distributed Systems
44
Smart Spaces
Smart School
Smart City
Smart Factory
โข Other Applications
Battlefields/Surveillance
Earthquake areas
Environmental Monitoring
Airport security
Emergency Response
Location Services
