简体   繁体   English

如何为 Ultimate Thread 组创建场景

[英]How to create scenario for Ultimate Thread group

I wish to create scenario for Ultimate Thread group for JMeter for 50 users, no time dependent constraint here as it's a Tableau based application.我希望为 50 个用户的 JMeter 的 Ultimate Thread 组创建场景,这里没有时间相关的限制,因为它是基于 Tableau 的应用程序。

There are total 46 APIs I've gathered from Swagger API documentation.我从 Swagger API 文档中收集了总共 46 个 API。

I've 5 user Id's which generate authorization token我有 5 个生成授权令牌的用户 ID

My client also not well versed in Load testing and they are expecting me to design the scenario for Load testing.我的客户也不精通负载测试,他们希望我为负载测试设计场景。

I'm also new to JMeter load testing, but I've created jmx script for all 46 API's with single user they all are working with success, and I'm not sure how to create Load scenario with Ultimate Thread group which will be most suitable for my project.我也是 JMeter 负载测试的新手,但是我已经为所有 46 个 API 和单个用户创建了 jmx 脚本,他们都在成功地工作,我不确定如何使用 Ultimate Thread 组创建负载场景,这将是最适合我的项目。

Someone told me using Ultimate Thread Group will be correct approach for creating the scenario, also I am the only tester in my scrum team, so I need to find out all suitable scenarios.有人告诉我使用 Ultimate Thread Group 将是创建场景的正确方法,而且我是我的 scrum 团队中唯一的测试人员,所以我需要找出所有合适的场景。

At end I need to share the report to Client, but the report generation done with JMeter is very complex, since I have 46 APIs and 50 users so the graphs are getting very complex, I also need some suggestion what will be the best and simple way to generate simple reports, which I can explain to client as well.最后我需要将报告分享给客户,但是使用 JMeter 完成的报告生成非常复杂,因为我有 46 个 API 和 50 个用户,所以图表变得非常复杂,我还需要一些建议,什么是最好和最简单的生成简单报告的方法,我也可以向客户解释。

I just created simple JMX script which I think would not be of any use until I create a good Load scenario and can generate a simpler report which will be understandable by me as well as by client.我刚刚创建了简单的 JMX 脚本,我认为在创建一个好的加载场景之前没有任何用处,并且可以生成一个更简单的报告,我和客户都可以理解。

  1. create a good Load scenario - there is only one criterion of a "good load scenario" - it must represent real life application usage as close as possible, better with 100% accuracy so hammering all API endpoints sequentially with 50 users is not something you should be doing, you should rather come up with real "user paths" which represent real users (or real applications) making use of your system under test create a good Load scenario - 只有一个“良好负载场景”的标准 - 它必须尽可能接近现实生活中的应用程序使用情况,更好地达到 100% 的准确度,因此您不应该使用 50 个用户依次敲击所有 API 端点这样做,您应该想出真正的“用户路径”,代表真正的用户(或真正的应用程序)使用您的测试系统

    Also it's not clear what type of test you're performing as there will be different setup for Load Test , Stress Test , Spike Test , etc. See Performance Testing vs. Load Testing vs. Stress Testing article to learn more about different performance testing sub-types and when they need to be applied此外,还不清楚您正在执行哪种类型的测试,因为负载测试压力测试峰值测试等会有不同的设置。请参阅性能测试与负载测试与压力测试文章以了解有关不同性能测试子的更多信息-类型以及何时需要应用它们

  2. generate a simpler report - if you apply these "user journeys" concept and put them under Transaction Controllers you will have much less data in charts. generate a simpler report - 如果您应用这些“用户旅程”概念并将它们放在事务控制器下,您将在图表中获得更少的数据。

    Also you can use Filter Results tool to get individual charts for individual APIs您还可以使用过滤结果工具来获取各个 API 的各个图表

    And last but not the least, if you generate HTML Reporting Dashboard you will get interactive charts where you will be able to turn on and off the data points you are interested/not interested in最后但并非最不重要的一点是,如果您生成HTML 报告仪表板,您将获得交互式图表,您可以在其中打开和关闭您感兴趣/不感兴趣的数据点

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