简体   繁体   English

蝗虫-Python负载测试。 没有提供文件; 那么上升的术语是什么意思?

[英]Locust - Python load testing. No documentation provided; so what do the ramping terms mean?

Locust gives the option to ramp up wherein you need to enter certain input details. 蝗虫提供了增加的选项,您需要输入某些输入详细信息。 I didn't find any documentation explaining the following terms whereas the other ones (not mentioned) seem self-explanatory: 我找不到任何解释以下术语的文档,而其他文档(未提及)似乎不言而喻:

Ramping 爬坡

Hatch stride
Precision (min value of hatch stride)
Calibration time (seconds)
Percentile (%)
Max percentile response time (ms)
Accepted fail ratio (%)

Thanks! 谢谢!

The ramping feature of Locust was experimental, and has been removed from the latest version. Locust的渐变功能是实验性的,已从最新版本中删除。

The idea with the ramping feature was that you would enter an acceptable failed request ratio ( Accepted fail ratio ), and a maximum acceptable response time ( Max percentile response time ) for a given percentile of users ( Percentile ), and then it would gradually ramp up the number of simultaneous users for as long as those limits weren't exceeded. 与斜坡功能的想法是,( 失败比接受 ),你将进入一个可以接受失败的请求比和可接受的最大响应时间( 最大百分比响应时间 )为用户给定的百分位数( 百分位 ),然后它会逐步地只要不超过这些限制,就可以同时增加用户数。 So the purpose was to automatically find out the number of users that a system could handle. 因此,目的是自动找出系统可以处理的用户数量。 However it never worked really well in practice and as Locust evolved the ramping feature was somewhat forgotten, and eventually removed. 但是,它在实践中从未真正有效地发挥作用,随着Locust的发展,斜坡功能被略微遗忘,最终被删除。

Hatch and Precision had to do with how fast new users would be spawned, and Calibration time was the time frame that that was used to measure the fail ratio and response time for a given percentile, before either spawning or killing more users. HatchPrecision必须与产生新用户的速度有关, 校准时间是在产生或杀死更多用户之前用于测量给定百分位数的失败率和响应时间的时间范围。

声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.

 
粤ICP备18138465号  © 2020-2024 STACKOOM.COM