New issue
Advanced search Search tips

Issue 725301 link

Starred by 1 user
Status: WontFix
Owner: ----
Closed: May 2017
Cc:
panicker@chromium.org
caseq@chromium.org
Components:
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 3
Type: Bug



Sign in to add a comment

Erroneous navigationStart events appear in traces with redirects

Project Member Reported by phulce@chromium.org, May 23 2017 
Chrome Version: 60.0.3103.0
OS: OS X

What steps will reproduce the problem?
(1) Create a trace using DevTools/Lighthouse/about://tracing for http://www.vkontakte.ru/ with Nexus 5X emulation enabled and the 'blink.user_timing' category enabled
(2) Open the trace in about://tracing

What is the expected result? A single navigationStart event for the main frame can be found that corresponds to the time reported by `window.performance.getEntriesByType('navigation')`

What happens instead? Multiple navigationStart events are present for the same frame (see attached trace file)
assets-0.trace.json
3.5 MB View Download

Comment 1 by phulce@chromium.org, May 23 2017

Description: Show this description

Comment 2 by panicker@chromium.org, May 23 2017

Components: Blink>Loader

Comment 3 by kenjibaheux@chromium.org, May 24 2017

Components: -Blink>Loader
Status: WontFix (was: Untriaged)
I'm seeing 2 navigation start events, one for what seems to be the about:blank page (or it could be a side effect of the redirect) and one for m.vkontakte.ru

Tracing reflects the internal workings of Chrome by design.
I think this is WAI.

Comment 4 by panicker@google.com, May 24 2017

Cc: caseq@chromium.org
Overloading the same trace event (with multiple values) is causing incorrectness for Devtools - which one should they use?
We could either have devtools use the first one OR make a unique event for "first actual nav start".
►

Sign in to add a comment