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

Issue 694145 link

Starred by 1 user

Issue metadata

Status: Archived
Owner: ----
Closed: Mar 2018
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: Linux
Pri: 2
Type: Compat



Sign in to add a comment

Repeated XHR requests leads to leak in extension

Reported by khym.cha...@gmail.com, Feb 20 2017

Issue description

UserAgent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/56.0.2924.87 Safari/537.36

Example URL:
http://www.pixiv.net/search.php?word=%E5%B5%90%E9%87%8E%E3%81%AE%E6%88%A6%E3%81%84

Steps to reproduce the problem:
1. Go to the given pixiv.net URL
2. Let page finish loading
3. Let page just sit there for a while

What is the expected behavior?
Page should finish loading in a reasonable amount of time, and the memory the page takes (according to Chrome's built-in task manager (Shift-Esc)) should remain constant over time as the page just sits there.

What went wrong?
1. The page makes repeated XHR GET requests for the same resources, over and over again.

2. The memory the page takes expands over time.  If left there for long enough it will die because it consumes too  much memory.

(NOTE: I think the above two points are two separate bugs, but I haven't been able to get them to reproduce separately)

Does it occur on multiple sites: No

Is it a problem with a plugin? No 

Did this work before? N/A 

Does this work in other browsers? N/A

Chrome version: 56.0.2924.87  Channel: stable
OS Version: Fedora 25
Flash Version: Shockwave Flash 24.0 r0

0. Attached is a bzip2'd tar file containing:
* pixiv.log: the console log of the XHR requests
* search.html: the HTML of the page displaying the problem
* ad_source.js: One of the JavaScript files issuing the XHR requests
* cedato_player_108.56_d.js: the other JavaScript file issuing the XHR requests

1. The problem might be specific to a single pixiv.net ad partner, so the page might have to be reloaded a few times before the problem shows.

2. The problems don't go away with either Shift-Ctrl-R or when all pixiv.net cookies are deleted with the EditThisCookie extension.

3. The problems *did* go away when I cleared the cache via chrome://settings/clearBrowserData, but this only worked once, and it came back when I did a Shift-Ctrl-R (hence why I think it might be restricted to just one of the ad partners of the site)

4. The problem goes away entirely if the page is loaded in a brand new user profile: there are no repeated XHRs, and the memory usage of the page remains stable.

5. If I start chrome with "--disable-extensions", the memory leak problem goes away, but the repeated XHR requests still remain.

6. If I start chrome with "--no-experiments", no difference is seen.

7. If I disable only the Stylish extension, the memory leak goes down a lot.

8. I have the following extensions installed and enabled:

<<<<<<
Vimium 1.57
1 Click Translator 8.4
EditThisCookie 1.4.1
eHistory 1.14
Extensions 1.0
Frame by Frame for YouTube™ 1.1.0
Image Properties Context Menu 0.7.6
Lazarus: Form Recovery 3.0.6
middle button new tab 0.9
Mute Tab Shortcuts 1.2.0
Mute Tabs By Url 2.5
Password Alert 1.23
Playing Favourites for Fallen London 0.3.1
Quick Javascript Switcher 1.4.1
Recently Closed Tabs 1.3.0.2
Session restore 0.6.2
Stylish - Custom themes for any website 1.7.3
Tabs to the front! 0.2.4
Vimium 1.57
Your Quality for YouTube™ 1.9
YouTube AutoFocus 0.5
>>>>>>>

9. I have a HAR (HTTP Archive) file of the network activity from when the page is displaying the bug, but I'm not attaching it here since it contains my cookies.  I will provide it to Chrome developers upon request.

 
FILES.tar.bz2
120 KB Download
Oh, almost forgot another point:

10. When I went to the "Sources" tab of the developers tool, it had bunches of copies of "ad_source.js".  Don't know if this is normal or not for a single URL that's XHR GET'd multiple times by the same page.
Labels: Needs-Triage-M56
Cc: krajshree@chromium.org
Components: UI>Browser
Labels: Needs-Feedback
Unable to reproduce the issue on Ubuntu 14.04 using chrome stable #56.0.2924.87 and latest dev #58.0.3013.3.

Attached a screen cast for reference.

Following are the steps followed to reproduce the issue.
------------
1. Navigated to url: http://www.pixiv.net/search.php?word=%E5%B5%90%E9%87%8E%E3%81%AE%E6%88%A6%E3%81%84
2. After the page loaded, opened the task manager using SHIFT + ESC.
3. Observed that the memory the page occupied (according to Chrome's built-in task manager (Shift-Esc)) remained constant over time as the page just sits there.

khym.chanur@ - Could you please check this issue on latest dev #58.0.3013.3 by creating a new profile without any apps and extensions and please let us know if the issue still persist or not.

Thanks...!!
694145.ogv
17.3 MB Download
1. Testing under 58.0.3013.3 with a new profile does not reproduce either problem (memory or repeated XHR requests)

2. Testing under 58.0.3013.3 with my old profile *does* reproduce both problems.
I can fix the problem by deleting the cookies files from my profile.  I no idea *which* cookie is causing the problem, though.
Labels: -Needs-Feedback
Thank you for providing feedback. removing "Needs-Feedback" label.
Cc: rbasuvula@chromium.org
Labels: Needs-Feedback
Tested in chrome # 56.0.2924.87 and canary #58.0.3029.0 on Ubuntu 14.04 and not able to reproduce the issue.

@ khym.chanur : As mentioned in comment #5 could you please confirm can we close the issue.

Thank you! 
It looks like I can reproduce the problem with a fresh profile, but it takes a while.  About a week ago I started out with a completely new ~/.config/google-chrome, then copied from the old ~/.config/google-chrome/Default to the new only the bookmarks, cookies, custom dictionary, shortcuts and visited links.   I re-did from scratch settings and extensions to match what I had before.

All of this caused the bug to go away, but today the bug is back.  So it seems to me that it takes almost a week of usage for the bug to manifest.

Oh, and blocking cookies from genieesspv.jp prevents the bug from manifesting.
Project Member

Comment 9 by sheriffbot@chromium.org, Mar 6 2017

Labels: -Needs-Feedback
Thank you for providing more feedback. Adding requester "rbasuvula@chromium.org" to the cc list and removing "Needs-Feedback" label.

For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot
Components: Platform>Extensions
Some one from extensions-dev team please take a look in to this issue.

Thank you!
Project Member

Comment 12 by sheriffbot@chromium.org, Mar 16 2018

Status: Archived (was: Unconfirmed)
Issue has not been modified or commented on in the last 365 days, please re-open or file a new bug if this is still an issue.

For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot

Sign in to add a comment