Sensitive data should be removed

from storage areas before another

subject accesses it

Object reuse

Processes should de-allocate

committed storage

Garbage Collection

Application Threats

Trap/Backdoors
Buffer Overflow
Covert Channels
Goal is to enforce an organizations
security policy and procedures in the
maintenance of CIA

Motivation and Study Techniques to help

you learn, remember, and pass your
technical exams!

Involves input to the system, the data being

processed, and the output from the system
Data Checks
Custom screens

CRCs

Backups

Visit us

Application Controls

Controls can be classed as

Detective

Hash's

CEH

www.mindcert.com

Corrective

Control reports

Goals

Surrogate program performs services

on behalf of another

Agent

Acts on behalf of principal but may

hide the principal

Applets

Produce a quality product that meets users needs

Stay within budget and time schedule

This is the phase where the programmers

and developers get down to business of
producing the product

proxy

Small apps
Downloaded from the web

Good understanding of the needs of security will help the

developers reduce the likelihood of buffer overflow and covert
channel vulnerabilities in the final product

Distributed System
Components

Assumes each step can be completed

without any effect from the later
stages of the development

Run in constrained space

Interpreted

Java

Simple Model

Assumes any reworking will not affect previous work

Multi platform
Establishes trust between client and
server with digital certificates

Define the concepts

Recognizes a need for developers to

modify early stages

Active X

Developers are limited to going back

one stage to rework
Large development teams to stop
parallel reworking

Project Initiation

Functional Design analysis and planning

Formal design

CISSP
More coming soon...

Application Controls

Validity checks

Define the requirements

Cisco

Preventative

Functional design review

Lifecycle

System life-cycle phases

Software development
Install

Evaluates product against spec

Verification

1976, rework to have all phases end

with a milestone and back references

Waterfall model with Verification and Validation

System Development

Installation

Test and audit

Not usually the case in the real world

Ending phase tied to a milestone

Models

Design specifications

Detail planning

Assumes a stage will finish at a specific time

Waterfall Model

Validation

Ensures real world requirement

Goes round in a spiral

Develop Plans

Continual product changes and fixes

Maintenance

Replace product with a new one

Define objectives

Spiral Model

Prototype and identify risk

Disposal

Final Development
Joint Analysis Development Model (JAD)
Hierarchical
Mesh
OO

Rapid Application Development (RAD)

Personnel away from the developers

should test the software

Types

Relational
Row in a relational model
A column in a data model

Element

Describes the database structure

Schema

Primary key

Attribute of one table that is the

primary key of another table

Foreign Key

Used as a security mechanism

Software maintenance
and Change Controls

Intersection of a row and column

Central repository for meta data and

data relationships

manage the change

Change Control

document everything
manage the upgrade

Release Control
Used to manage evolving changes to
software products

Track Versions
Issues new products

Should conform to BS 7799

Component to be changed

Configuration item

View

CISSP

Cell

Application and
Systems Develop

Version

Configuration Management

Recorded state of the Configuration item

Collection of component configuration

Configuration
Definitions

Building

Data Dictionary

The number of rows in a relationship

Build List

Cardinality

Subtopic
Subtopic

Software Library

Select

Benefits of a formal
software process

Project
union

Three phases

Jargon

Virtual relationship to display specifics

Also called an element

Software Development

Candidate Key

Field that links all the data in a row

The requirement for change

Request Control

Attribute

Data in a cell

Identifier that is unique to the record

Keeps testing objective

Unit testing should be addresses when

modules designed

Tuple

Primitives

Difference

Subtopic

Improved quality
Reduced life-cycle time
More accurate scheduling and meeting of
milestones

product
Level 1

Join
Intersection
Divide
Controls database access

Level 2

Non-Primitives

Defined from Join, project, and select

Important operation

Appears as a virtual table with settings

that the user can view

Initiating

Relational Operations

View

Software Capability
Maturity Model (CMM)

Quality of the finished product is a

component of the quality of the
development process

Five Levels

Ensures that attributes in a table rely

only on the primary key

Level 3

Level 4

Repeatable

Defined

Managing

Eliminates repeating groups

Eliminates redundant data

Data Normalization

Eliminates attributes not dependent on the primary

key

Level 5

Database concepts
and security issues

Select
Update

Potentially capable of being

more reliable and reduces the
possible propagation of program
change errors

Delete
insert
Access Privileges
Access Privileges

SQL

Grant
Revoke

Classes

Security is provided in relational

databases through views

Items

Virtual relation that combines

information from other relations

Put data into strict containers and

limit access control

Encapsulation

OOP Concepts
Security Issues

Fundamental Characteristics

Containers

Cell suppression

Implement a view partition scheme

insert bogus information in the database
A record of a higher level of security
holds different data to the same
record of a lower one

Data hiding

Polymorphism

Different objects can react to identical

messages in different ways

OO Systems
Polyinstantiation
Inheritance

Using a view control to hide specific cells

happen, but cannot see

Objects perform work by sending

messages to other objects

Aggregation

Inference

Data called attributes

Often called black box functions

Messages

Is ability of users to infer or deduce

info about data at sensitivity levels for
which they do not have access

A link is called an inference channel

Called methods

Objects contain procedures

Objects

The DBMS can be compromised by

circumventing the normal security controls
Act of obtaining information of higher
sensitivity by combining information
from lower levels of sensitivity

These tell the system how to make objects, the

process of creating an object using the
directions in a class is called "instantiation"

Allows an object to be copied and

populated with different data

Subclasses inherit settings

All predefined types are objects

All user defined types are objects

Anti Aggregation and Inference measures

All operations are performed by

sending messages to objects

Noise

Distributed Systems

Polyinstantiation

Distributed Component Object Model (DCOM)

Made available to users over a network

Repository of heterogeneous databases that is

available for users to make queries

Object Request Brokers (ORB)

Data is normalized
Redundant data is removed

Common Object Request Broker

Architecture (CORBA)

Data Warehouse

ORBs are middle-ware

Establishes a client server relationship
between objects

Can be applied to audit logs to find

system anomalies
Objective is to find relationships that were
unknown up until now in the warehouse
Data about Data
Metadata is not stored in the
warehouse but in the Data mart
Database system for developers
Stores all data structures used by an application

Data Mining

Data Warehousing

Metadata

Data Mart
Data Dictionary

CISSP Application and Systems Developement.mmap - 15/05/2009 - Andrew Mason

Optimizing

Chaotic
quality is unpredictable
PM exists
No formal method
Formal processes in place

Product improvement
Process improvement
Continuous process improvement
Bugetized