Friday, April 22, 2016

Top 20 Question you should ask when selecting a Logical Data Warehouse

If you are evaluating or researching a Logical Data Warehouse (LDW), it is likely that you are looking for a way to eliminate all the overhead and implementation time of your current Data Warehouse or to reduce the proliferation of databases and all the overhead associated with those. You may be looking to use it for one or more scenarios, for example:

  • Support Business Intelligence and Analytics with clean, fresh data consolidated live from multiple sources
  • Standardize usage of data throughout the enterprise with Data as a Service (DaaS)
  • Generate actionable and repeatable Master Data definitions
Logical Data Warehouse           

Classic Data Warehouse         

The following 20 questions will help you to make sure you won’t need to augment with additional tools once you are into your project.

Top 20 Questions to Ask When Selecting a LDW platform


Basic:

  1. Is your LDW agile? Because your requirements are constantly changing, and you need to be able to make changes in a matter of seconds or minutes
  2. Can the LDW connect directly to every source you need? You don’t want to have to invent custom programming to force feed your LDW. That defeats the purpose.
  3. Are all the consumption methods you need supported? ODBC, JDBC, OData, SharePoint BCS, SOAP, REST, Ado.Net, others
  4. Can you configure in-memory and on-disk caching and refresh for selected sources that do not need to be refreshed at each query? This improves performance and alleviates load on the source system. In many situations, you really don’t need all of the source data updated on every query if it doesn’t change much or often. The best platforms will have full ETL capabilities.
Ease of Use and Configurability:

  1. Can you configure, test, deploy, and monitor from a single platform?
  2. Does it have design time debugging? You don’t want to keep going in and out of the platform to test.
  3. Can you re-use all components, validations, rules, etc.?
  4. Is there numbered versioning on configurations, including the who/what and time stamp?
  5. Is there auto packaging and self-hosting of all services?
Enterprise Readiness:

  1. Can you do data validation on source data?
  2. Is there a full transformation engine for complex business rules about combining data?
  3. Is it available on-premise and IPaaS?
  4. Is there an Execution audit trail?
  5. Can a virtual data model be a source to another virtual data model?
  6. Is there write-back to sources with transaction rollback?
  7. Is there end user authentication for full CRUD (Create, Read, Update, Delete) for “actionable” dashboards and such?
  8. Does it handle streaming data?
  9. Are there multiple points for performance tuning?
  10. Does the Platform have individual user logins specifying the permissions for each? For example, maybe a DBA can view and modify a transformation, but not create one.
  11. Is there definitive data lineage available including validation rules, federation rules, etc.?
Want to learn more about the Logical Data Warehouse?  Download the whitepaper.

2 comments:

  1. This comment has been removed by a blog administrator.

    ReplyDelete
  2. Really interesting content which provided me the required information.
    Interesting visit:-
    warehouse management software systems

    ReplyDelete