Geek Sync I The Importance of Data Model Change Management

Geek Sync I The Importance of Data Model Change Management

• 1.
  The Importance ofData Model Change Management March 8, 2017 Joy Ruff Product Marketing Manager Joyce.Ruff@idera.com
• 2.
  2 Agenda  Enterprise datatrends  Development methodologies  Communicating through data models  Considerations for change management  Sprint-based modeling activities  Summary  Q&A
• 3.
  3 Enterprise data trends Increasingvolumes, velocity, and variety of Enterprise Data 30% - 50% year/year growth Decreasing % of enterprise data which is effectively utilized 5% of all Enterprise data fully utilized Increased risk from data misunderstanding and non-compliance $600bn/annual cost for data clean-up in U.S.
• 4.
  4 Evolving Database Ecosystems Volume, Velocity, Variety Keepingpace with the rapid growth of data, change and compliance Agile Development Cycles Maximizing IT Infrastructure ComplianceLimited Resources Data Professionals Need the Right Tools
• 5.
  5 Waterfall vs Agile DataModeling
• 6.
  6 Data model usage& understanding 13% 3% 16% 19% 31% 18% 0% 5% 10% 15% 20% 25% 30% 35% We don’t use data models Other Our data team does most data models but developers also build… Our database administrators own data modeling Developers develop their own data models We have a data modeling team that is responsible for data models Completely understand 20% Understand somewhat 60% Don’t understand 17% I don’t know 3% 87% What is your organization’s approach to data modeling? How well does your organization’s technology leadership team understand the value of using data models?
• 7.
  7
• 8.
  8 Why we needdata models: Much more than a picture  Full Specification • Logical • Physical  Descriptive metadata • Names • Definitions (data dictionary) • Notes  Implementation characteristics • Data types • Keys • Indexes • Views  Business rules • Relationships (referential constraints) • Value Restrictions (constraints)  Security classifications + rules  Governance metadata • Master Data Management classes • Data quality classifications • Retention policies
• 9.
  9 Benefits of datamodeling  Design • Manage redundancy • Integrate and rationalize • Increase quality  Use & Maintain • Increase discoverability • Improve comprehension • Data dictionaries • Business glossaries
• 10.
  10 The Need forCommon Understanding
• 11.
  11 Apply meaning withbusiness glossaries  Maximize understanding of the core business concepts and terminology of the organization  Minimize misuse of data due to inaccurate understanding of the business concepts and terms  Improve alignment of the business organization with the technology assets (and technology organization)  Maximize the accuracy of the results to searches for business concepts, and associated knowledge
• 12.
  12 Data model changemanagement considerations  Needs to work with any workflow style – not just sprint-based  Fine-grained check-in and check-out capability  Method to associate model changes to requirements and list them in a change management control center  Audit trail of changes made – what was done and why, to demonstrate compliance for data governance  Ability to compare models to databases and other models, and identify changes that need to be merged into the source or target  Capability to create branches from a model baseline and merge them back in or roll back to restore a previous release  Ability to generate the necessary DDL code to implement the desired changes into the database
• 13.
  13 Agile data modelingconsiderations  Primary focus is enablement of the team • Can not be perceived as an obstacle/gatekeeper  Iterative work style • Managing changes during sprints • Implementing database changes with DDL  Collaboration is paramount • Cross-project focus • Enterprise data perspective  Traceability – what changed and why • Data lineage can show change impacts • Audit reporting for data governance
• 14.
  14 Managing changes duringagile sprints
• 15.
  15 Start of sprintpreparation  Participate fully in sprint planning  Ensure there is a “Named Release” as of completion of previous sprint • Always have a baseline for compare/merge!  Submodels • Structure by relevant topic/subject area • At story level if necessary to facilitate communication • Roll up to parent level submodels
• 16.
  16 In-sprint activities  Modelerfully engaged in daily stand-up meetings  Model change workflow • Model each change, associating with appropriate task/user story • Generate incremental DDL script(s) and post • Use a robust script naming convention, particularly if utilizing automated build systems  Different work approaches • Some designs will be originated “pushed” by data modeler • Others may be “pulled” from developer “sandbox” • Analyze, amend and “push” back out • Compare/merge and redesign as appropriate • Ensure developer uses the officially sanctioned script • Use submodels for audience specific perspective • Use data model design patterns  Maintain the discipline!
• 17.
  17 End of sprintwrap-up  Create “Named Release” at end of Sprint • Serves as baseline for start of next sprint • Serves as baseline for comparison at ANY later point  Create delta DDL script by using compare/merge • Based on Named Release from end of the previous sprint  Create full database DDL script • Can be used to easily create “sandbox” databases quickly  Ensure the model(s) have been published  Participate fully in sprint planning and retrospectives • Lessons learned • Celebrate the successes
• 18.
  18 Summary  Important factorsfor effective data model change management • Working style • Agile (sprint-based) or Waterfall • Collaborative rather than gatekeeper • Consistency in communication • Data dictionaries • Business glossaries • Traceability • Track why changes are made, not just what • Correlate with development changes • Implement model version control • Provides audit trail for compliance / governance
• 19.
  19 Thanks! Any questions? You canfind me at: joyce.ruff@idera.com @jfruff