The model process of Master Kimball is: from and the whole process is bottom-up. This structure is also known as Data Mart Bus Architecture or Bus Architecture. The core part of the Kimball model is dimensional modeling. If you understand this part well, the understanding of the Kimball model can basically do more with less. This article focuses on dimensional modeling. 2. Application environment 1. Definitions Dimensional modeling is a data structure technology whose main purpose is to help users perform large-scale and efficient data queries.
Applicable The above definition illustrates that the mobile number list purpose of dimensional modeling is to optimize the database to query data faster. The dimensional model is designed to read, summarize, and analyze the numerical information in the data warehouse, for example: want to see the summary value, quantile value, etc., this feature is more suitable for the data warehouse environment. In the transaction database (OLTP), this design method is not suitable, because the transaction database has the following characteristics.
Good for fast updating and inserting data Locking must be minimized Only suitable for a small subset of data queries need to be standardized Let's take a look at the characteristics of the warehouse environment: Suitable for faster query data Locking is optional Query for large amounts of data Inserts and updates are not necessary So dimensional modeling needs to be applied in the data warehouse environment. 3. Explanation of terms Before introducing the overall modeling process, some terms need to be introduced in detail, so that the following content will be easier to read.