All users were logged out of Bugzilla on October 13th, 2018

Create a list of target pages & documents that represent different transaction types

RESOLVED FIXED

Status

RESOLVED FIXED
3 years ago
3 years ago

People

(Reporter: groovecoder, Unassigned)

Tracking

({in-triage})

Details

(URL)

(Reporter)

Description

3 years ago
Add the page types & URLs here:

https://etherpad.mozilla.org/MDN-load-testing

Note: Can use GA to find the pages with the most traffic
(Reporter)

Updated

3 years ago
Blocks: 1186081
(Reporter)

Updated

3 years ago
Blocks: 1186091
(Reporter)

Updated

3 years ago
No longer blocks: 1186081
(Reporter)

Updated

3 years ago
Keywords: in-triage
(Reporter)

Comment 1

3 years ago
Per GA, this should be a good list to start using for performance regression testing.

1. en-US/
2. en-US/docs/Mozilla/Projects/Midas/Security_preferences : high traffic, >200 contributors, no KS
3. en-US/docs/Web/JavaScript : high traffic; ~175 contributors, sidebar, 3 KS macros
4. en-US/docs/Web/JavaScript/Reference/Global_Objects/Array/forEach: high traffic, ~60 contributors, 6 parent docs, extensive sidebar, table of contents, 31 KS macros
5. en-US/docs/Web/Guide/CSS/Media_queries : high traffic, ~50 contributors, TOC, long content, 8 languages, 117 KS macro calls
6. en-US/Learn/Getting_started_with_the_web/JavaScript_basics : high traffic, zone page
7. en-US/demos/
8. en-US/Learn high traffic, zone landing page
9. ru/ - highest-traffic non-English home page
10. zh-CN/docs/Web/JavaScript - highest-traffic non-English doc page

Page traffic:

1. 1.25%
2. 0.87%
3. 0.67%
4. 0.58%
5. 0.55%
6. 0.23%
7. 0.16%
8. 0.14%
9. 0.16%
10. 0.09%

Of the sampled traffic, that makes the proportionate traffic amounts:

1. 26.5%
2. 18.5%
3. 14.3%
4. 12.3%
5. 11.7%
6. 4.9%
7. 3.4%
8. 3%
9. 3.4%
10. 2%

So, I'll start working on a performance regression smoke test that uses these urls and sends these url requests in the right proportion.
Status: NEW → RESOLVED
Last Resolved: 3 years ago
Resolution: --- → FIXED
You need to log in before you can comment on or make changes to this bug.