web| Skills | data | database | Performance Our design goal is to make the test program in the various platform algorithm as far as possible consistent.
While keeping the code easy to understand and develop quickly.
If performance is not a major concern, then a variety of different technologies can lead to very different speed differences.
For example, Microsoft rewrote the Nile benchmark we used to demonstrate optimization, and got different application logic and database design, and tested twice times as much on the same hardware platform.
The following are tips for developing database applications using scripting languages:
1, put the application logic into the database.
The less the page needs to be displayed, the faster it will run. Therefore, try to use stored procedures to reduce the middle layer of scripting logic.
2, optimize the database as far as possible.
Make full use of cursor technology (especially forward cursors), reduce transaction isolation, utilize read-only result sets, make full use of precompiled statements and boundary variables.
3, reduce network traffic between the Web server and the database.
By using a batch feed array, the method of manipulating the data in memory is also written to the database in a batch way.
4, do not store variables and user settings to the application server's memory.
Store them in cookies or databases to reduce application server lock and CPU loading content
5, the use of buffering to make the dynamic page into static.
Refresh the page periodically using a timer.
6, reduce the number of HTML and scripting code switching between each switch, how much will affect performance.
7, to ensure that. html or. htm files contain only HTML so that the scripting engine does not have to analyze documents that are not necessary to analyze.
