Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Fix load shape worker in headless. #1539

Merged
merged 2 commits into from Aug 21, 2020
Merged

Conversation

cyberw
Copy link
Collaborator

@cyberw cyberw commented Aug 21, 2020

No description provided.

@cyberw
Copy link
Collaborator Author

cyberw commented Aug 21, 2020

ping @max-rocket-internet

@cyberw
Copy link
Collaborator Author

cyberw commented Aug 21, 2020

I should add a test for this as well...

@cyberw
Copy link
Collaborator Author

cyberw commented Aug 21, 2020

btw, are you on locust slack channel?

Copy link
Contributor

@max-rocket-internet max-rocket-internet left a comment

Yeah I've never used headless TBH, we just use API/UI 🙂

@cyberw
Copy link
Collaborator Author

cyberw commented Aug 21, 2020

Cool, then you'll get all the UI bugs from now ;P

Btw, have you checked out the Timescale+Grafana stuff in locust-plugins? In my own (very partial :) opinion, it is much better than using the UI :)

@codecov
Copy link

codecov bot commented Aug 21, 2020

Codecov Report

Merging #1539 into master will increase coverage by 0.28%.
The diff coverage is 0.00%.

Impacted file tree graph

@@            Coverage Diff             @@
##           master    #1539      +/-   ##
==========================================
+ Coverage   82.47%   82.76%   +0.28%     
==========================================
  Files          28       28              
  Lines        2562     2564       +2     
  Branches      389      390       +1     
==========================================
+ Hits         2113     2122       +9     
+ Misses        348      346       -2     
+ Partials      101       96       -5     
Impacted Files Coverage Δ
locust/main.py 21.05% <0.00%> (+0.69%) ⬆️
locust/runners.py 84.57% <0.00%> (-0.20%) ⬇️
locust/user/task.py 97.83% <0.00%> (+1.62%) ⬆️
locust/clients.py 95.09% <0.00%> (+4.90%) ⬆️

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 709f189...7b5589a. Read the comment docs.

@cyberw cyberw merged commit 2abfdda into master Aug 21, 2020
2 of 3 checks passed
@max-rocket-internet
Copy link
Contributor

max-rocket-internet commented Aug 24, 2020

Cool, then you'll get all the UI bugs from now ;P

😅

Btw, have you checked out the Timescale+Grafana stuff in locust-plugins?

I had a quick look just now. We don't use the locust UI for evaluating the test results or looking at the performance of the application under test. If we are running a non-automated test then we just use it to see if the test is running, if it's healthy (exceptions etc) and check worker count.

We are big Prometheus users and have custom dashboards for our applications (and for our locust k8s infrastructure). We use these for evaluating the test results as they are also what we use for live environments so the results are comparable.

For automated load tests we use Spinnaker pipeline to scale up locust and application infrastructure, start the load test via locust API, monitor exceptions and error rates via the locust API for a set duration, then stop the test afterwards (or if exceptions and error rates are too high).

@cyberw
Copy link
Collaborator Author

cyberw commented Aug 24, 2020

Btw, have you checked out the Timescale+Grafana stuff in locust-plugins?

I had a quick look just now. We don't use the locust UI for evaluating the test results or looking at the performance of the application under test. If we are running a non-automated test then we just use it to see if the test is running, if it's healthy (exceptions etc) and check worker count.

We are big Prometheus users and have custom dashboards for our applications (and for our locust k8s infrastructure). We use these for evaluating the test results as they are also what we use for live environments so the results are comparable.

Nice! Havent really used Prometheus that much myself, but it looks really nice.

For automated load tests we use Spinnaker pipeline to scale up locust and application infrastructure, start the load test via locust API, monitor exceptions and error rates via the locust API for a set duration, then stop the test afterwards (or if exceptions and error rates are too high).

Cool stuff. I wanted Klarna to use Spinnaker when I worked there but I dont think I ever convinced anyone (when I was there I think they were building something similar in house :-/ )

One of the nice things about timescale/the logging plugin is I get exact metrics for every request, logging for the entire test run (and trends for aggregated results as well).

But I guess if you have good monitoring on the system side then you may not need that (as I'm more of a consultant performance tester, I dont always have access/knowledge of the underlying system's metrics)

@cyberw cyberw deleted the fix-shape-functionality-for-headless branch Mar 22, 2022
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

2 participants