New issue
Advanced search Search tips
Note: Color blocks (like or ) mean that a user may not be available. Tooltip shows the reason.

Issue 704135 link

Starred by 2 users

Issue metadata

Status: Assigned
Owner:
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: Android
Pri: 3
Type: Bug



Sign in to add a comment

Measure the renderer process startup performance for the case of creating a second webview after the renderer has exited

Project Member Reported by tobiasjs@chromium.org, Mar 22 2017

Issue description

We measure startup performance currently, but we don't capture the case where all WebViews are destroyed (leading to the renderer exiting) and then create a new WebView.

Although doing this is a memory saving, it may also represent a latency regression for applications.

We should track this also to ensure that we don't regress from where we are currently.
 
We should also measure the cost of creating a webview when the renderer already exists. Anecdotally this is about 1/10 of the initial webview cost in single process mode.
Cc: nednguyen@chromium.org
Components: Tests>Telemetry
Cc: -nednguyen@chromium.org nedngu...@google.com
Components: -Tests>Telemetry Speed>Benchmarks
Project Member

Comment 5 by sheriffbot@chromium.org, Apr 12 2018

Labels: Hotlist-Recharge-Cold
Status: Untriaged (was: Available)
This issue has been Available for over a year. If it's no longer important or seems unlikely to be fixed, please consider closing it out. If it is important, please re-triage the issue.

Sorry for the inconvenience if the bug really should have been left as Available.

For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot
Owner: changwan@chromium.org
Status: Assigned (was: Untriaged)
Changwan, it seems like - given your current focus - this should be assigned to you.
Cc: pasko@chromium.org
Cc: -nedngu...@google.com nednguyen@chromium.org
Cc: paulmiller@chromium.org torne@chromium.org
Yes, I noticed this while measuring startup time locally. Usually, startup time was 4x or 5x longer when that happens - I'm not sure how it's reflected in our perfs now, but I think setting such a separate metric would be a good start.

Sign in to add a comment