Some basic knowledge of OLAP and OLTP

Oltp:on_line Transaction processing online transaction processing

Olap:on_line Analytical processing online Analytical processing

OLTP, as its name suggests, focuses on business processes. OLAP is specially designed to support complex analysis operations, focusing on decision support for decision makers and senior managers, it is possible to quickly and flexibly perform complex query processing of large amounts of data, and provide the query results in an intuitive form.

The main differences between OLTP and OLAP are as follows:

(1), the user and system to be oriented: OLTP is customer-oriented and is handled or queried by an employee or customer. OLAP is to the market, by managers, supervisors and analysts for data analysis and decision-making.

(2), data content: OLTP systems manage current data, which is often trivial and difficult to use for decision making. OLAP systems manage large amounts of historical data, provide aggregation and aggregation mechanisms, and store and manage information at different granularity levels, which makes data appropriate for decision analysis.

(3), database design: Typically, OLTP uses ER model and application oriented database design, while OLAP systems usually adopt star mode or snowflake mode and theme-oriented database design.

(4), view: The OLTP system focuses on the current data of an enterprise or department, and does not involve historical data or data from different organizations. In contrast, OLAP systems often span multiple versions of an enterprise's database schema, and OLAP systems also process information from different organizations, integrated by multiple data sources.

(5), Access mode: OLTP system access is mainly composed of short atomic transactions, this system requires concurrent control and recovery mechanism. Most of the access to OLAP systems is read-only, most of which are complex queries.

(6), metrics: OLTP focuses on day-to-day real-life operations, so with transactional throughput as a measure, OLAP is measured by query throughput and response time.

The above preliminary identification of these two concepts, later for a mouthful of the concept of people can be slightly defensive.