request canceled while waiting for connection (Client.Timeout exceeded while awaiting headers)

[zbitname]

Expected behavior

Pull and run container.

Actual behavior

Docker return error with message:
(HTTP code 500) server error - {"message":"Get https://registry-1.docker.io/v2/: net/http: request canceled while waiting for connection (Client.Timeout exceeded while awaiting headers)"}
at pull image.

Information about the Issue

Platform: Win10 x64
Docker version 17.06.0-ce, build 02c1d87
I don't use proxies and VPN. Speed of internet connection is ~90Mb/s.

Steps to reproduce the behavior

I reproduce this issue on creation container with "mongo". Percent of reproduce this issue is ~95%, sometimes everything goes well.

[FrenchBen]

Please make sure that you're on a stable connection and use a solid DNS such as Google DNS.

[saibhaskerraju]

Go to Docker settings > network > DNS server . change from automaic to fixed ( default is 8.8.8.8 )

[SiddiquiImran]

moby/moby#22635 (comment)

[fastinetserver]

I had a similar issue trying to login to my own docker repo.
Here is how i've managed to resolve it:
https://stackoverflow.com/questions/46822391/docker-toolbox-tutorial-client-timeout-exceeded-while-awaiting-headers/48064026#48064026

[matheuslemke]

My problem was missing proxy config:

nano /etc/systemd/system/kitematic.service.d/proxy.conf

[Service]
Environment=http_proxy=http://host:port

systemctl restart snapd

[Tony427]

Go to Docker settings > network > DNS server . change from automaic to fixed ( default is 8.8.8.8 )

This is god damn works !!

[adamfortuno]

I had this exact issue with the "jenkins/jenkins:lts" container. I'm running...

docker pull jenkins/jenkins:lts

I get the following error:

Running the 10.0.17763.0 build of Windows 10 (64-bit).

I resolved by restarting the Docker service.

[juanmigutierrez]

Go to Docker settings > network > DNS server . change from automaic to fixed ( default is 8.8.8.8 )

Works !

[cvladan]

Go to Docker settings > network > DNS server . change from automaic to fixed ( default is 8.8.8.8 )

Works !

I also confirm this is the solution.

[zllegol]

Works

[mohdumaira]

Go to Docker settings > network > DNS server . change from automaic to fixed ( default is 8.8.8.8 )

Thanks for the solution. I was stuck for half a day on this. It works.

[stella-lu]

Does anyone know how to do this now? I tried going to "Preferences" under Docker, but I can't find anything about network or DNS server so I don't know how to change the DNS server settings.

I am using Docker Desktop, Docker version 18.09.2 on MacOS Mojave version 10.14.5

[nclsjstnn]

Does anyone know how to do this now? I tried going to "Preferences" under Docker, but I can't find anything about network or DNS server so I don't know how to change the DNS server settings.

I am using Docker Desktop, Docker version 18.09.2 on MacOS Mojave version 10.14.5

i had to go preferences to proxies > manual configuration proxy > and set web server (http) to 8.8.8.8

[diegoldev]

Installing Docker update 2.1.0.1 fix the problem for me.

[hsmiranda]

i'm resolved this problem with setting proxy variables correct, oc or kubectl utilize proxy variables for connect to api even on localhost.

[minato7]

Faced similar issue within corporate network while trying to access internal registry from Windows host. Resolved it by removing the Proxy server configuration.

I had set proxy in Web Server section due to which only hub.docker.com was only accessible. If necessary to access registry with Proxy then add it.