Michael Watson

1
Team 3

• Michael Watson
• Alicia Stevenson
• Tamya Stallings
• Steve Gibson

2
Topics of Discussion

• Summary Data Michael Watson
• Security Alicia Stevenson
• Encryption Tamya Stallings
• Budget and Cost Justification Steve Gibson

3
Summary Data in the Data Warehouse/ODS Environment

• There is a split between the operational and
  informational environments
• Operational transaction and day-to-day
  processing
• Informational analytical data

4
Information Systems Architecture
5
Detailed/Summary Data

• Detailed data has existed for a long time and
  forms the backbone of operational processing
• Summary data has came about in recent years with
  the implementation of data warehouses

6
Framework of Summary Data

• Whether summary data is dynamic or static,
• Whether the usage of summary data is predictable
  or unpredictable, and
• Whether the summarization results in a low level
  of summarization or a high level of summarization.

7
Framework of Summary Data
8
Implementation Dimensions

• Whether the data is permanently stored or stored
  on a temporary basis, and
• Whether a single summary snapshot of data is
  produced or whether a series of time sequenced
  summary snapshots are produced.

9
Elements of Summarization
10
Monitoring/Sweeping Summary Data
11
Platforms the Summary Data Resides On
12
Security
13
Security

• Gartner Study
• 80 not secure by 2005
• Security is often ignored
• Security Example Teksouth
• access and analysis system with 128-bit
  encryption

14
Corporate Information Factory (CIF) Security
15
Two Threats to Security

• External
• Those outside the enterprise that would misuse
  information found inside the CIF
• Through the Internet
• Internal
• Someone inside the organization who is familiar
  with the structure and the workings of the CIF,
  and unauthorized access is made
• Seldom come over the Internet

16
Corporate Information Factory

• Components
• Data marts
• DSS applications
• Alternative storage
• Data mining warehouses
• ODS

17
Data Marts

• The data warehouse is the center of the CIF
• Lower level of security
• Large central facility belonging to the central
  IT organization
• Protects data through standard system facilities
  like network monitors, firewalls, and standard
  operating system facilities

• The data marts are the departmental components of
  the CIF
• Higher level of security
• Owned and managed by departments themselves
• The department can use firewalls, network
  security, DBMS security and operating system
  security.

18
Data Mining

• Least amount of security
• Derived information
• Token based technology

19
ODS

• Need for security and protection
• ODS data is used for online decision making
• Highest level of security throughout the entire
  CIF
• Central IT organization

20
Encryption

• Definition scrambling a message so no one can
  read it except for the person who its intended
• Plaintext vs. ciphertext
• More profound level of security
• Very time consuming to manage

21
Data Warehouse Encryption

• Data cant be encrypted in variable lengths
• Must be decrypted in the same format
• Domain of data must be preserved
• Example if the data contains a number then it
  must be encrypted as a number
• Vital to ensuring the system doesnt collapse
• Protects data from dumps

22
Unique requirements

• Ability to encrypt some columns and not others
• Part of the data could be encrypted one way with
  the other half of the data encrypted a different
  way (interleaving affect)

23
Encryption restrictions

• Keys are not encrypted
• Columns that use a WHERE clause are not encrypted
• Foreign key relationships arent encrypted
• Data used in standard database management system
  processing are not encrypted

24
Encryption abuses

• Example Employee, manager, and data clerk all
  need same data
• Encryption doesnt distinguish between users
• Degrades performance and lessens the data
  availability to users
• Requires complicated key administration

• Performance degradation
• One or two elements may not cause degradation
• Entire table of elements performance degradation
  is considerable

25
Encryption abuses (cont.)

• Good practice
• Laptops
• Encrypt all sensitive data
• Protects is stolen

• Main priority
• Select encryption
• Various organizational steps for protecting
  physical site and media

26

• Budget is important because it shows where the
  true priorities are.
• It shows where the money is being spent in the
  data warehouse
• Two types of cost
• 1) one-time
• 2) continuous

27
One time costs

• Hardware
• Disk storage
• Software
• Integration and transformation interface

• Processor
• Network communication
• System management tools
• Metadata creation and population

28
Continuous Costs

• Refreshment of data
• End user training
• Maintenance and update of data
• Administration costs
• Periodic verification of the conformance to the
  enterprise data model
• Servicing data mart request
• Reorganization and restructuring of data
• Archiving of data

29
These items cost differently depending on many
factors

• Size of the organization
• Amount of historical data
• Level of detail
• Sophistication of the end user
• Company is in a competitive market
• How fast the data warehouse grows
• Number of data marts stemming from it
• Centralized or distributed
• Constructed manually or automated

30
One time costs

• Hardware
• Disk storage ...30
• Processor costs ... 20
• Network communication costs ...10
• Software
• DBMS...10
• Access/analysis tools ... 6
• Systems management tools
• Activity monitor 2
• Data monitor .2
• Integration and transformation interface
  creation.15
• Metadata creation and population..5

31
Continuous costs

• Refreshment of the data warehouse data 55
• Maintenance of infrastructure... 3
• End user training .6
• Data warehouse administration
• Conformance to the enterprise data model2
• Servicing data mart requests for data21
• Capacity planning .. 1
• Monitoring of activity and data7
• Occasional reorganization/restructuring of data
  1
• Archiving of data ..1
• Summary table usage analysis 2
• Security administration ..1

32
Hard to justify cost

• The benefits cannot be seen until after the data
  warehouse is built
• Many factors contribute to the benefits, not just
  the data warehouse

33
Compare data warehouse to legacy1 report 10 in
data warehouse 1000 in legacy environment

• Computer resources from the legacy environment
  cost more the warehouse
• Legacy environment is sensitive to response time
  unlike warehouse
• Much data in legacy environment that can get in
  the way
• Internal structure of data warehouse was built
  for reports, legacy was not
• Technology in data warehouse was optimized for
  reports, legacy was not
• Data in warehouse is integrated, legacys data is
  not

34
Compare continued

• Data warehouse has metadata to locate data
  quickly
• Metadata lets DSS analyst know if report was
  already made
• More historical data in data warehouse to use
• More summary data in data warehouse
• Quicker to create one in data warehouse
• Data warehouse is ideal for discovery mode and
  changing a report is easy

35
Controlling costs

• Use an automation tool for creation and
  maintenance of the code used to move data from
  legacy to data warehouse
• Store the warehouse on a combination of disk
  storage and near line storage
• Build the data warehouse iteratively- spiral
  method best
• Use log journals and tapes as a source of
  refreshment

36
Controlling costs can't

• Use a consultant
• Use a data model as the basis of the design
• Capacity analysis done at the beginning of
  development
• Research what vendors to buy software and
  hardware from
• Purge data occasionally
• Never build a data mart first

37
Mikes Question

• What are the three important dimensions of
  summary data?

38
Alicias Question

• What are the different components of the
• Corporate Information Factory?

39
Tamyas Question

• What are the four restrictions for encryption
  data in a data warehouse?

40
Steves Question

• Q What are two types of cost and give at
  least two examples of each