0% found this document useful (0 votes)
12 views

SIP17 Course Sumary

Uploaded by

Hưng B
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
12 views

SIP17 Course Sumary

Uploaded by

Hưng B
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 22

System Integration

Mini Case Studies © 2010

Course Summary and Wrap Up

Shawn A. Butler, Ph.D.


Senior Lecturer, Executive Education Program
Institute for Software Research
Carnegie Mellon University
Objectives
 Review some of the major points of the
course
 Summarize lessons learned

© 2010 CMU-ISR 2
Why are you integrating?
 Adaptable systems and processes
 Streamlined business processes
 Management information
 Support for electronic commerce
 Integrated security
 Replaceable components
 Reliable and recoverable systems
 Economies of scale

© 2010 CMU-ISR 3
Integration Heuristics
 Simplify, Simplify, Simplify
 Don’t assume that the original statement
of the problem is necessarily the best, or
even the right one
 Build and maintain options as long as
possible in the design and implementation
of complex systems. You will need them
 In partitioning, choose the elements so
that they are as independent as possible;
that is, elements with low external
complexity and high internal complexity
© 2010 CMU-ISR 4
SDLC Phases
 Conception – What are the goals and
objectives of the integrated system? What
is the vision?
 Requirements Analysis – What are the
requirements of the integrated system?
 Specification – What are the system
detailed specifications?
 Design – What are the components and
how will they communicate?

© 2010 CMU-ISR 5
SDLC Phases cont’d
 Implementation – Integrate!
 Testing – testing and more testing
 Training and Support – Make sure the
users know how to properly use what they
paid for!
 Maintenance – Fix bugs and make changes
as necessary
 Retirement – This almost never happens!

© 2010 CMU-ISR 6
3 Models of Integration
 Presentation Integration – information is
integrated through the user interface
component
 Data Integration – information is
integrated through a middleware layer that
has business intelligence
 Functional Integration – information is
integrated among databases or data
sources

© 2010 CMU-ISR 7
Common Services
 Naming Service
• Every entity sees the entire system as the same set of
objects
• Define interfaces to entities
• Users know how to find other system entities

 Security Service
• All entities (principles) are authenticated with some
degree of reliability
• Access to resources are managed based on defined
policies
• Activities can be logged to ensure accountability and
recovery

 Reliability – The ability of the system to provide


necessary functionality despite unexpected events
© 2010 CMU-ISR 8
Consistency
 Autonomous data sources are in a consistent
state when a user can access data at any source
and it will be the same
 After an update, how long before the system will
be in a consistent state?
 What is the requirement? Immediately, within a
few minutes, hours, or next day?
 Consistency requirements must be determined in
distributed systems based on user requirements!
 The faster the consistency must be achieved, the
more expensive the system

© 2010 CMU-ISR 9
Why Do We Care about
Consistency Issues?
 Modern Transaction Processing products
ensure information consistency
• Not always – Not all system objects covered
• You may want to relax consistency
requirements to improve performance
• Cross domain transactions create integration
problems
 Different transaction processing products
 Different assumptions about consistency

© 2010 CMU-ISR 10
Two Applications with
Overlapping Functionality
 Applications may have very similar, but slightly
different functionality
 Application design has tightly coupled sub-
components – Not easy to break apart
 Each sub-component may make assumptions
about other subcomponents
 Redundancy in functionality adds maintenance
overhead
 Sub-component not designed to be reused
 Especially problematic with redundant services

© 2010 CMU-ISR 11
Possible Solutions for
Overlapping Functionality
 Choose one application and extend functionality
to meet total functionality
 Refactor one (or both) of the applications so that
specific functionality can be extracted and then
integrated
 Create API’s to increase sub-component
independence
 Wrap components and create new interfaces or
clients
 Ignore functionality

© 2010 CMU-ISR 12
Integration Guidelines
 Application Coupling – Minimize
dependencies
• Tightly coupled applications have known and
unknown assumptions
• Applications can evolve independently without
problems

 Intrusiveness – Minimize changes in each


application
• Changes are often necessary
• Tradeoff between intrusiveness and best
integration design

© 2010 CMU-ISR 13
Integration Styles
 File Transfer
 Shared Databases
 Remote Procedure Invocation
 Messaging Systems

© 2010 CMU-ISR 14
Data Integration Mistakes
 Creating yet another database
 Waiting for the data analyst to finish developing
the perfect schema
 Implementing the perfectly normalized schema
 Assuming the data exists as described in the
documentation
 Testing without an sufficient set of real data
 Assuming that one site is a good representation of
data at all sites

© 2010 CMU-ISR 15
Relational Databases
 Database normalization rules are designed to
prevent anomalies and inconsistencies in
databases
 Database normalization rules, strictly applied,
may introduce inefficiencies in database design
 All databases have various schemas that work
well for the their application, but don’t combine
well into an efficient schema for all applications
 Database design is always a tradeoff among
application performance optimization

© 2010 CMU-ISR 16
Security Integration
 Authentication, access control, and
auditing are the fundamentals of system
security
 Integration of security services is difficult
and takes considerable planning
 Integration of security services may
introduce more risk than the risk of each
component
 Fundamental element of security is risk

© 2010 CMU-ISR 17
Middleware
 Messaging is one of the most popular middleware
technologies for integrating disparate applications
• Asynchronous communication, but can be designed to do
synchronous communication
• Different types of MOM
 Point to point message queuing
 Broadcasting

 CORBA not as popular, but still out there


 Lots of middleware options
 Transaction Processing Monitors ensures ACID
properties in heterogeneous environments

© 2010 CMU-ISR 18
Accessing the System
 On the network
• Local Area Networks
• Standing Virtual Private Networks
• Wide Area Networks

 Remotely
• Remote Desktop Protocol (RDP)
• Terminal Services
• Virtual Private Networks

 DNS, routers, and Active Directory are key


to sharing information across a network
© 2010 CMU-ISR 19
Clients – Fat or Thin?
 Provide adequate interactive performance
 Ensure the integrity of system updates
 Minimize risk of sensitive information
exposure
 Minimize network activity
 Enable disconnected user activity

© 2010 CMU-ISR 20
Enterprise Client Devices
 Specialized devices are becoming an
essential part of the enterprise solution
• Cell phones/PDA’s
• Mobile media devices

 Protection of these devices is a significant


challenge
 User interfaces change among different
devices
 Wireless connections are expanding

© 2010 CMU-ISR 21
Summary
 There are many aspects to integrating a system
 System engineers have to understand a broad
spectrum of hardware and software technologies
 Good system engineers are good because they
have made many mistakes, or have seen many
mistakes
 Your ability to anticipate changes, see how the
system fits together, and pay attention to detail
are good skills for a system engineer
 When on a project, use your journals to
remember the things that can go wrong!

© 2010 CMU-ISR 22

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy