admin管理员组

文章数量:1146251

For instance, login user-action. Say User-action 'login' has total of 5 APIs. (One API actually performs the login and rest of the APIs perform some other actions e.g. load home page, load some other meaningful data on that page etc). My question is about the performance statistics report:

  1. should we focus on individual API level stats (sequential execution) or
  2. should our performance report focus more on user level actions (how much time the login user-action took, regardless of how much time each API took individually)?

For instance, login user-action. Say User-action 'login' has total of 5 APIs. (One API actually performs the login and rest of the APIs perform some other actions e.g. load home page, load some other meaningful data on that page etc). My question is about the performance statistics report:

  1. should we focus on individual API level stats (sequential execution) or
  2. should our performance report focus more on user level actions (how much time the login user-action took, regardless of how much time each API took individually)?
Share Improve this question edited 2 days ago DarkBee 15.7k8 gold badges69 silver badges111 bronze badges asked 2 days ago Faisal IqbalFaisal Iqbal 411 silver badge2 bronze badges
Add a comment  | 

1 Answer 1

Reset to default 0

Well-behaved JMeter test should act exactly like a real user using a real browser.

Real browsers execute AJAX requests using concurrent thread pool so my expectation is that you should mimic what you see in "Waterfall" tab of your browser developer tools, to wit:

  1. First request to login
  2. Followed by X concurrent requests which load home page and meaningful data and these requests need to be parallel, not sequential. Take a look at i.e. Parallel Controller which can simulate browser's network footprint in this scenario.

本文标签: jmeterShould we use individual request (API) level performance stats or useraction levelStack Overflow