Charteris Integration Tools

1
Charteris Integration Tools

• The Cost-Effective Approach to Data Integration
• Robert.worden_at_charteris.com

2
The Problem - Data Integration

• Most organisations have overlapping data in
  hundreds of databases and message formats.
• They cannot integrate this data, compare it,
  report on it, or keep it consistent across
  systems
• The business and IT costs are huge
• For governments, the problem is even bigger
• SOA and Cloud Computing do little to solve the
  problem they just make it more pressing

3
Approaches to Data Integration

• Two approaches are used in practice
• Mapping and automatic data translation
• Hand-coding of translations and interfaces
• Hand-coding dominates.
• Most mapping tools have limited capability, and
  have to be supplemented by hand-coding.
• Why?

4
Types of Mapping
S2S
Data Structure 1
Data Structure 2
(RDBMS, XML)
S2M
S2M
S2M
Class Model 1
Class Model 2
(UML)
M2M
S2M Structure-to-Model Mappings (similarly
for S2S, M2M)
Most Mapping Tools do Structure-to-Structure
Mapping (S2S)
5
Why Structure-to-Model Mapping is Necessary

• A UML class model expresses the logical
  information content of data
• To translate from one data structure to another,
  you must convey all types of information in the
  class model (objects, properties, associations)
• Structure-to-structure mappings do not let you
  map the associations (try it!)
• So you cannot generate reliable or complete data
  transformations from structure-to-structure
  mappings

6
Structure-to-Model Mappings are Cost-Effective
30 structure-to-structure Mappings, and 30
system interfaces
6 structure-to-model mappings
System 1
System 6
UML Class Model
System 5
System 2
Structure-to-model mappings solve the N-squared
cost explosion
30 data translations are generated from 6 sets of
mappings
System 4
System 3
7
Features of the Charteris Tools

• Mapping Editor, to create S2M mappings
• Comparative Query Tool, to query data from
  different systems side by side
• Automated Data Transformation between different
  system formats
• Round-Trip Accuracy of data transformations is
  guaranteed and tested.
• Integrate with Eclipse tools Java, XML,
  database, Web, UML, EMF,Talend, Borland

8
Benefits of the Mapping-Based Approach

• Rapid exploration of data across many systems,
  side by side (expose problems early)
• Build interfaces and migrate data with much less
  effort
• Avoid the N2 problem for many systems
• Guaranteed Integrity of Interfaces (e.g. round
  trip consistency)
• Automated self-test of data translations

Lower cost, higher quality of data integration
and migration
9
Application Types Using Data Integration

• Data Migration
• Business Intelligence and Reporting
• Data Quality
• Real-time Application Integration
• Customer Data Integration
• Enterprise Architecture/Data Architecture
• SOA Data integration
• Cloud Data Integration

10
Next Steps

• Download and install the tools
• Follow a Guided Tour, to see what the tools can
  do.
• Choose a pilot application area (with databases,
  message formats, business benefits)
• Carry out a short Proof of Concept
• Create or capture a small UML logical model
• Build some mappings
• Use the mappings to query data
• Use the mappings to translate data between system
  formats
• Plan your first live project.

Charteris can assist you with steps 3,4,5