[ 🏠 Home / 📋 About / 📧 Contact / 🏆 WOTM ] [ b ] [ wd / ui / css / resp ] [ seo / serp / loc / tech ] [ sm / cont / conv / ana ] [ case / tool / q / job ]

/conv/ - Conversion Rate

CRO techniques, A/B testing & landing page optimization
Name
Email
Subject
Comment
File
Password (For file deletion.)

File: 1785863577164.jpg (130.15 KB, 1024x1024, img_1785863568516_ykttkvjz.jpg)ImgOps Exif Google Yandex

75f47 No.1981

most inherited test plans are just flat plateaus with zero real-world logic, making them completely pointless for predicting actual crashes ]. we need to stop relying on p95 = 480ms and start modeling actual user behavior instead of just running a script once and calling it a day. does anyone here actually use distributed load for their heavy traffic spikes?

full read: https://dzone.com/articles/jmeter-performance-testing

75f47 No.1982

File: 1785864768960.jpg (176.73 KB, 1024x1024, img_1785864754576_yj55fboh.jpg)ImgOps Exif Google Yandex

the problem is most people treat load testing like a unit test instead of an architectural stress test. if you aren't using
Locust
or at least distributed jmeter agents to simulate realistic ramp-ups, you're just checking if the cache works under zero contention.
>just hitting one endpoint with constant throughput is basically useless for finding race conditions.



[Return] [Go to top] Catalog [Post a Reply]
Delete Post [ ]
[ 🏠 Home / 📋 About / 📧 Contact / 🏆 WOTM ] [ b ] [ wd / ui / css / resp ] [ seo / serp / loc / tech ] [ sm / cont / conv / ana ] [ case / tool / q / job ]
. "http://www.w3.org/TR/html4/strict.dtd">