Data now assumes a common role in controlling software logic
Data now assumes a common role in controlling software logic, yet often sidesteps the careful controls which are in place for source code changes. The business need to adapt quickly often conflicts with the careful controls in place to manage source code changes.
Business users sometimes realize they can bypass these controls by changing configuration data. Changing configuration data can cause applications to take a different courses of action based on new settings. What was once hard-coded and carefully controlled can now be dynamically reconfigured. It is considered chaotic to change code ‘in the fly’ hence this rarely occurs on-purpose. Changing configuration data ‘on the fly’ however is often all too easy.
Changing code on-the-fly is considered chaotic, but changing configuration data on-the-fly is all too easy
Even when authority to change configuration data is restricted to just IT staff the problem still persists. Without implementing similar careful controls on configuration data changes as we have for source code changes, developers and testers can easily introduce undesirable consequences, either by accident or deliberately.
Extends software configuration management to configuration data
Data configuration management must:
Easily identify and maintain a list of configuration data.
Including definition of files, fields and data areas.
Package configuration data into versions.
Perform configuration data transfers to development, to test and to production the same way you would with source code.
Automatically maintain an audit trail.
See which configuration data elements were changed, see before and after images of the change, when changed and who changed.
If a critical problem occurs implement a rapid rollback of configuration data to a known correct version.