Data optimization considerations
Another consideration when preparing your data for tabular models is the data refresh options available. Typically, data is imported into your tabular model similar to the process we used with multidimensional models. Imported data is loaded into memory and optimized by the VertiPaq engine. This involves a high level of compression, including columnar data storage techniques. The functions of compression and memory combine to create an optimized model with performance. Here are some key considerations when using data refresh:
- Refresh frequency: The data is only as fresh as the last import. If the data source has been updated recently, the data may be out of sync. This is less of an issue when you are loading data from a data warehouse. The data warehouse is typically loaded in batches as well. If you match your refreshes to the batch loads, your data will be consistent with the data warehouse. If you have chosen to use the transactional database...
