Skip to content

Data Mart Access - OverviewΒΆ

Data Mart access means accessing snapshot data. When reading On Line, data is fetched as of now from several sources and it might be necessary to include sub-Selects and function calls in the query to retrieve the requested information.

Data Mart can be seen as performing read of On Line data at a specific point in time and then evaluating and storing the information in a snapshot table. Reading from snapshot tables has several advantages:

  • The stored data more or less represents the raw data related to a Dimensions or Fact. Only row levels security filtering and data that is dependent on things like the user or preferred language needs to be added on top of the raw data.
  • Data Mart related snapshot tables can easily be indexed from the Aurena client, meaning that indexes can be added individually for each customer with respect to data volumes and use cases.
  • Read performance will in general be much better than if reading from the On Line source.
  • For entities that support incremental load, the update of associated snapshot tables can be done incrementally and leads to faster refresh than having to read large data sources fully or conditionally (e.g. load all data where year_of_date > 2018).

A snapshot table can be one of the following:

  1. Materialized View - This is an Oracle specific snapshot table that has a SELECT statement as the source expression. A Materialized View can support incremental load if a number of conditions are met but IFS has decided to not use that functionality. This means that all Dimensions and Facts that has a Materialized View that serves as the snapshot table, only support full load (every refresh requires a full load).
  2. Incremental Table - IFS Cloud has a framework that supports incremental load of Dimensions and Facts and that is using an ordinary Oracle table as the snapshot table. Development of incremental support is done by the Product Teams within R&D. So far this support has only been developed for Facts, typically since Facts represent transactional sources with potentially large number of transactions.

Use the following links to learn more about: