At present, data calculation has become one of the essential high-frequency operations in the analysis work, and in the context of the development of big data, application performance is often related to the success or failure of the project, and has become the product technical parameter that everyone pays most attention to. Then let's first analyze the reason why the calculation of the BI analysis table is very slow, and then prescribe the right medicine!
The three main reasons for the slow calculation of business intelligence BI analysis reports are as follows:
1. The amount of data is too large.
Business intelligence BI analysis reports need to process a large amount of data, which may come from multiple data sources, including internal data and external data. When the amount of data is too large, the calculation speed of the business intelligence BI analysis report will become very slow. When dealing with large amounts of data, the system needs to spend more time for data reading, storage, processing and transmission, resulting in slower calculation speed.
2. Complex query statements
Business intelligence BI analysis reports usually require complex queries and data processing. These query statements may involve multiple data sources and multiple tables, and multiple screening , sorting, aggregation and other operations are required. If the query statement is poorly written or the query statement is too complex, it will lead to slow calculation speed. At this time, it is necessary to optimize the query statement to minimize the load of data transmission and data processing, so as to improve the calculation speed.
3. Insufficient system resources
Business intelligence BI analysis reports need to occupy a lot of system resources, including CPU, memory, disk space, etc. If the system resources are insufficient, it will cause the calculation speed to slow down. At this time, system resources need to be optimized or upgraded to improve system performance and computing speed.
Yixin Huachen has been deeply involved in the field of business intelligence for more than ten years, and has a deep understanding of the pain points of customer needs. Therefore, Yixin ABI regards comprehensive performance optimization and the pursuit of the ultimate "fast" as the top priority. It has been greatly optimized and improved, mainly for fast calculation, fast rendering, and fast concurrency. Next, Xiaoyi will introduce and share with you in combination with the performance test report.
1. Optimization of loading performance of large data volume analysis tables With the company's business expanding, many companies will involve various systems, accumulate a large amount of data, and form analysis tables
with huge data volumes one after another. Whenever we need to analyze the table When previewing, it takes time to wait for the presentation of the calculation results. In order to solve this problem, Yixin ABI has carried out in-depth optimization of page rendering, and implanted intelligent analysis in the system. The background can automatically judge the analysis table with large data volume, and choose the method of full optimization or batch optimization to load and optimize the analysis table , and at the same time, when scrolling to view the report data, the page will be smoother and smoother without jamming. Through comparative tests, we obtained the following data results. When batch optimization is used, the loading speed can be increased by up to 15 times.
2. Cache calculation
Aiming at the data analysis function most frequently used by users, the new version implements analysis based on high-speed topics (report templates, WPS/Word analysis reports, Kanban, impromptu reports, slide reports are all supported), compared with Because of the way of using the physical theme table, the calculation speed has been significantly improved. This part is optimized by using Yixin's self-developed cache technology, and there is no need to additionally install the MPP database, and the operation is simple to check. After use, the calculation speed of the report is increased by 13 times compared with the original one, and the greater the amount of analyzed data, the more obvious the increase in calculation speed.
We can personally experience this optimization effect when using fixed table analysis, agile kanban, and ad hoc reporting functions. We simulated user scenarios and conducted comparative tests on computing performance for different data volumes, as shown in the following figure:
3. Multi-user concurrent login optimization
For large corporate groups, multi-user concurrent login is the norm. The new version of Yixin ABI has greatly improved the transmission data, which has greatly improved the login response time, and the internationalized login interface is also Optimized for synchronization.
According to the performance report, when 50 users log in concurrently, the response speed increases from 11.3s to about 4.5s, which is 1 to 2 times faster than the original speed. And the more concurrent accounts, the more obvious the speed advantage.
In addition to efficient data processing capabilities and high-performance data caching technology , Yixin ABI also has the advantages of intelligent analysis engine and convenient integrated development. As an all-round data analysis platform, Yixin ABI has opened up the entire data application link from data access, to data modeling and processing, to data analysis and mining, and can meet various complex analysis needs in enterprise operations. Help enterprises achieve efficient digital transformation.