New issue
Advanced search Search tips

Issue 914729 link

Starred by 1 user

Issue metadata

Status: Closed
Owner: ----
Closed: Jan 14
Components:
EstimatedDays: ----
NextAction: ----
OS: Windows
Pri: 2
Type: Bug



Sign in to add a comment

Chrome does not show that download failed if host stopped responding mid download

Reported by artjom.k...@gmail.com, Dec 13

Issue description

UserAgent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/70.0.3538.110 Safari/537.36

Example URL:
It was local application

Steps to reproduce the problem:
1. Start downloading large file like pdf with specified header Content-Length so you can see how much is left to download (1 / 100mb)
2. Kill host that is transferring file stream 
3.  Since HTTP socket is closed Chrome instantly shows that download is finished like it was successful.

What is the expected behavior?
If chrome did not download full content-length of the file and the socket is closed Chrome should show that download has failed. 

What went wrong?
I am sending Content-Length header in response so chrome knows what is the file size, but even if it is not fully downloaded. Chrome should check if full Content-Length has been downloaded when the socket is closed and show status failed if it failed to download the whole file.

Did this work before? N/A 

Chrome version: 70.0.3538.110  Channel: n/a
OS Version: 10.0
Flash Version: 

Firefox and Internet explorer have this behavior and it would make sense for chrome to have it too.
 
Components: UI>Browser>Downloads
Labels: Needs-Triage-M70
Labels: Needs-Feedback
Could you attach a NetLog of this happening? Thanks!
https://www.chromium.org/for-testers/providing-network-details

(I set up a quick test server and was not able to reproduce this.)
ping: reporter, could you please answer the question in comment #3?

Status: Closed (was: Unconfirmed)
Closing since no feedback has been provided. If this issue still exists, please file a new bug, including the net-log of this happening, as mentioned in comment #3.

Sign in to add a comment