At present, we are doing a project of separating the front end from the back end. The front end uses Node framework and express template and ejs.
The project has been started for a period of time, and the basic functions of the front page have been completed at present, so I did a pressure test locally. First of all, I don’t know whether to strive for or be professional in doing this local pressure test, and I also ask experts in this field to give advice on how to do the pressure test correctly.
After the program is launched, it may only run on a single machine. At present, the optimization method I know is to put NGINX in front of NODE, and then NGINX does some optimization processing. These can also be found on the Internet. Node has few optimization programs on the Internet, and there are also few specific optimization schemes.
My machine Asus R510VC i5-3230M 12G memory 480GSSD
Post the results of your local pressure test
ab -n 1000 -c 100 http://127.0.0.1:3000/ Server Software: Server Hostname: 127.0.0.1 Server Port: 3000 Document Path: / Document Length: 24797 bytes Concurrency Level: 100 Time taken for tests: 21.421 seconds Complete requests: 1000 Failed requests: 0 Total transferred: 25123620 bytes HTML transferred: 24797000 bytes Requests per second: 46.68 [#/sec] (mean) Time per request: 2142.130 [ms] (mean) Time per request: 21.421 [ms] (mean, across all concurrent requests) Transfer rate: 1145.35 [Kbytes/sec] received Connection Times (ms) Min mean[ plus /-sd] median max Connect: 0 1 2.8 0 16 Processing: 203 2041 775.0 1837 4996 Waiting: 203 2037 774.8 1834 4964 Total: 203 2041 775.0 1839 4996 Percentage of the requests served within a certain time (ms) 50% 1839 66% 1893 75% 1938 80% 1971 90% 3303 95% 4056 98% 4652 99% 4745 100% 4996 (longest request)
I hope you can give me some advice, and it is better to say it in more detail, that is, regarding the results of the pressure test, such as the average website throughput to what level is qualified, to what level is cow force, or compare dishes, etc
Thank you all!
Generally speaking, it mainly looks at rps(Requests per second) and needs to be tested under different values of C and N. Make a change graph and test many times, taking the mean value, because the A/B Testing result is not very stable, it is meaningless for you to give a test result like this.