I'm trying to understand the relationship between Essbase and traditional data warehousing. I thought it was just an alternative for data warehousing based on certain types of data. Is this true? Is it just another source that can be used in addition to or in replacement of a traditional data warehouse and what is an example of using Essbase instead of a data warehouse? What would be an example of needing both sources? Thanks all.
-LM
The relationship is that a company would typically use the DW as a source of data for Essbase. You wouldn't use Essbase as DW, though. Essbase would be used specifically to provide specific analytics based on a subset of DW data. For example, consider a DW that contains all data in an organization: sales records, HR information, operational data, and so on. There may be and often is a very specific subset of that data (say, financial data by location) that a company would want to load from the DW (often by way of a bespoke data mart or data lake) to a particular Essbase "cube". Whereas the DW may have financial data down to the transaction or journal entry level, it would be much more common to just load summarized information to the cube. For example, you may need to sum the financial data by account, year, period, and location and then load that into the cube to provide analytics ("slice and dice"/"ad hoc reporting").
You may be wondering what Essbase brings to the table versus just sourcing the data from the DW itself. Essbase (or a similar OLAP engine) lets you arrange data into dimensions with hierarchies, supports "write back", and is fast. Write back allows users to submit their own data back to the cube. This is typically used to facilitate budgeting and forecasting activities. Ad hoc reporting gives users (typically finance users that have been trained how to use Essbase via its plugin for Excel) the ability to create reports that retrieve data very quickly (as opposed to how long an equivalent query on the DW/DM might take), as well as offering the advanced reporting capabilities offered by having the dimensions and hierarchies arranged in a manner that is conducive to the users and business. For instance, while you may typically load an Essbase cube with totals by month, it's very common for the Time dimension in an Essbase cube to aggregate or "roll up" to Quarters, which in turn aggregate up to a total for the year. Similar aggregations are designed for business entities (arrange by state, organization, region, etc), financial accounts (arrange by lines on financial documents, P&L, etc.), and more.