Bug 259160 – [ macOS iOS ] http/tests/privateClickMeasurement/triggering-event-with-attribution-source-through-fetch-keepalive.html is a flaky text failure

• Component: New Bugs
• Status: NEW

[Robert Jenner]

http/tests/privateClickMeasurement/triggering-event-with-attribution-source-through-fetch-keepalive.html

is a flaky text failure on macOs and iOS. It fails more frequently on Mac, than it does on iOS. But it does appear to be flaky on both. 

HISTORY:
https://results.webkit.org/?suite=layout-tests&test=http%2Ftests%2FprivateClickMeasurement%2Ftriggering-event-with-attribution-source-through-fetch-keepalive.html

TEXT DIFF:
+CONSOLE MESSAGE: Origin http://localhost:8000 is not allowed by Access-Control-Allow-Origin. Status code: 404
+CONSOLE MESSAGE: Fetch API cannot load https://localhost:8443/.well-known/private-click-measurement/trigger-attribution/12?attributionSource=https://127.0.0.1 due to access control checks.
+CONSOLE MESSAGE: Unhandled Promise Rejection: TypeError: Load failed
 CONSOLE MESSAGE: [Private Click Measurement] Triggering event was not accepted because the conversion data could not be parsed or was higher than the allowed maximum of 15.
 CONSOLE MESSAGE: Origin http://localhost:8000 is not allowed by Access-Control-Allow-Origin. Status code: 404
 CONSOLE MESSAGE: Fetch API cannot load https://localhost:8443/.well-known/private-click-measurement/trigger-attribution/Dummy?attributionSource=https://127.0.0.1 due to access control checks.


DIFF URL:
https://build.webkit.org/results/Apple-Monterey-Debug-WK2-Tests/265988@main%20(8187)/http/tests/privateClickMeasurement/triggering-event-with-attribution-source-through-fetch-keepalive-pretty-diff.html

[Radar WebKit Bug Importer]

<rdar://problem/112155451>

[EWS]

Test gardening commit 266005@main (6a50dd8d2c2e): <https://commits.webkit.org/266005@main>

Reviewed commits have been landed. Closing PR #15794 and removing active labels.

[Robert Jenner]

Interestingly enough I also see it failing this way on Ventura:

TEXT DIFF:
 Tests triggering of private click measurement attribution with same-site triggering event request.
 
 
-Attributed Private Click Measurements:
+Unattributed Private Click Measurements:
 WebCore::PrivateClickMeasurement 1
 Source site: 127.0.0.1
 Attribute on site: localhost
 Source ID: 3
-Attribution trigger data: 12
-Attribution priority: 0
-Attribution earliest time to send: Within 24-48 hours
-Destination token: Not set
+No attribution trigger data.
 Application bundle identifier: testBundleID

DIFF URL:
https://build.webkit.org/results/Apple-Ventura-Release-AppleSilicon-WK2-Tests/265957@main%20(3970)/http/tests/privateClickMeasurement/triggering-event-with-attribution-source-through-fetch-keepalive-pretty-diff.html


And I was also able to trivially reproduce this failure at Ventura Release ToT running the test as follows:


run-webkit-tests http/tests/privateClickMeasurement/triggering-event-with-attribution-source-through-fetch-keepalive.html --iterations 100


It only reproduced 2 out of 100 times when I ran in iterations.

[EWS]

Test gardening commit 265870.50@safari-7616-branch (081893a11bbe): <https://commits.webkit.org/265870.50@safari-7616-branch>

Reviewed commits have been landed. Closing PR #685 and removing active labels.

[Ben Schwartz]

This test has the same diff log and similar history as the following test:

http/tests/privateClickMeasurement/triggering-event-with-attribution-source-through-fetch-keepalive-ephemeral.html

Marking new test expectations.

[Ben Schwartz]

Upon further inspection, the issue persists across these three tests:

http/tests/privateClickMeasurement/triggering-event-with-attribution-source-through-fetch-keepalive.html
http/tests/privateClickMeasurement/triggering-event-with-attribution-source-through-fetch-keepalive-ephemeral.html
http/tests/privateClickMeasurement/attribution-conversion-through-fetch-keepalive.html
http/tests/privateClickMeasurement/attribution-conversion-through-image-redirect-ephemeral.html

Marking the appropriate test expectations for the three tests other than this one.

[Ben Schwartz]

After further review, the two diffs Robert mentioned are likely two separate issues. I have decided to file a new bug for the tests I mentioned instead of modifying the test expectations under this bug.

New bug link: https://webkit.org/b/260052