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

Issue 893958 link

Starred by 12 users

Issue metadata

Status: Duplicate
Merged: issue 895476
Owner: ----
Closed: Oct 24
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: Windows
Pri: 2
Type: Bug


Show other hotlists

Hotlists containing this issue:
Chrome-ES


Sign in to add a comment

[User Feedback - Stable] 500%+ increase in ERR_SSL_VERSION_INTERFERENCE errors in Chrome 69.100

Project Member Reported by melodychu@chromium.org, Oct 10

Issue description

Chrome Version: 69.0.3497.100 - it appears the increase correlates with this version
OS: Win 10, Win 8, ect

Many of the reports mention Gmail or Facebook:
"Cannot open Gmail from chrome.  Was using it Ok earlier this morning.  Other Google apps (e.g. photos and calendar) are working OK. Can access gmail from phone (nexus 4). Have tried clearing browsing data."

"For 4 days I have not been able to access gmail or facebook through the google chrome browser."

What happens instead?
Link to reports: http://shortn/_h2zeIrKxJh
 
Is there a way to ask these users to file bugs and, in particular, ask about their network and any antivirus, etc? This is likely correlated to switching from the draft version of TLS 1.3 to the final one, which should have been a completely trivial change. Some middleware must be actively misbehaving to be sensitive to this.
Oh wait, this is M69, not M70? That's bizarre. We didn't change anything in M69.
(Though the same question about middleware applies.)
Cc: susan.boorgula@chromium.org davidben@chromium.org
Labels: Triaged-ET Needs-Feedback
Tested this issue on Windows 10 on the reported version 69.0.3497.100 and unable to reproduce the issue.
Could login to Gmail and facebook on 69.0.3497.100 and no issues are observed.
Attached is the screen cast for reference.

This issue looks similar to issue 894046. Hence CC'ing davidben@ and requesting to check and help in further triaging of this issue.

Thanks..
893958.mp4
3.4 MB View Download
Right, that's to be expected, as it would be a problem with the network.

melodychu, do you know if comment #1 is possible? I'm not familiar with that feedback system.
We've gotten some other reports on crbug, also on M69. Given the lack of correlation with anything on our end and that M69 hasn't changed its TLS 1.3 for several releases now, it's most likely some AV vendor screwed up. Hopefully one of the crbug reports will be enlightening.
Cc: viswa.karala@chromium.org
 Issue 895079  has been merged into this issue.
Cc: vamshi.kommuri@chromium.org
 Issue 895121  has been merged into this issue.
 Issue 894947  has been merged into this issue.
 Issue 889805  has been merged into this issue.
 Issue 895150  has been merged into this issue.
 Issue 895148  has been merged into this issue.
Cc: svaldez@chromium.org
 Issue 894724  has been merged into this issue.
Please don't merge things into this bug. User reports are *extremely* valuable because, unlike this ConOps bug, we can actually ask users what happened. I'm going to unmerge them until we hear back.

It currently looks like Kaspersky shipped a bad update. They will need to fix this on their end.
We've filed  issue #895476  to dup the Kaspersky entries into. At this point I think we can fairly solidly say that's what's going on, though we have no real way of confirming it for the ConOps reports.

We've reached out to Kaspersky. Given how quickly this happened, Kaspersky likely has an auto-update system and so they should be able to remedy this fairly smoothly.
everything appears to be back to normal. Thank you for assistance
but problem now attaching pdf document to gmail with chrome
doing ok now
It appears like the increase has disappeared since Kaspersky issuing a patch late last week.

Should this bug get closed?
Mergedinto: 895476
Status: Duplicate (was: Untriaged)
I think we can merge this into the main Kaspersky bug. The drop confirms that was indeed the cause.

Sign in to add a comment