Pull is still not working for me but in my case it’s related to the hosts company proxy settings not exposed to the docker engine. The fix is work in progress: #13144
I managed to get pull to work by using the VPN compatibility mode but it only allows so much data to be donwloaded. So if I try to install anything within a container it will download so far and then fail.
I had a few interfaces with the name “vEthernet (DockerNAT)” installed, but only one was connected but with no IP configured. After I configrued the ip 10.0.75.1 everything was fine!
we noticed that under some circumstances some old “vEthernet (DockerNAT)” are lingering around. We haven’t figured out what causes this and are not able to reproduce it, but had updated the torubleshoot page a while back to provide instruction suggesting to delete all of them.
Beta13 and Beat14 also have some significant changes to the networking setup which should make this mode stable as well.
Specifying the DNS in docker GUI worked for me. I put an internal DNS since we have local registry with private domains. Also if I disable the Ethernet adapter it works as well. I have Ethernet and Wireless adapters connected to the network.
Tried disabling IPv6 for the vEthernet (DockerNAT) switch and also specifying Google DNS, but no dice. Where can I find the log files that people are posting here? No idea where these are, and I can’t find any documentation for it.
Also, here are the versions of the tools I’m using:
PS C:\Users\evan.cox> docker --version
Docker version 1.12.0-rc3, build 91e29e8, experimental
PS C:\Users\evan.cox> docker-compose --version
docker-compose version 1.8.0-rc1, build 9bf6bc6
PS C:\Users\evan.cox> docker-machine --version
docker-machine.exe version 0.8.0-rc1, build fffa6c9
Windows 10 Pro 10.0.10586 Build 10586
Also, we run GlobalProtect for VPN off-site, but right now it’s disconnected and I don’t think I’m behind a proxy.
Well, I spoke too soon, Docker just notified me of a new release,1.12.0-rc4-beta19, one of the bug fixes was the truncated TCP dns response fix. With this new build, docker run hello-world works! No mention of the beta 19 release on the website yet, but it’s rolling out.
Had the same problem on a colleagues machine. Unable to pull images.
Did as some have suggested however it didn’t work.
Disabled IPv6
Set the on the DocketNAT interface to 8.8.8.8 and 8.8.4.4
I then checked to make sure NAT was working ok. Turns out the NAT object had disappeared. You can check this by using the “Get-NetNat” cmdlet in powershell.
Once I created it manually I was able to pull images again.
Hope this is helpful to anyone else that’s having this problem lately. Will hopefully get to submit a proper bug report sometime today with my findings when I get some time.
Still same issue for me with the latest version (1.12.0-rc4-beta20) on a fresh install of Windows 10 Pro. Docker won’t pull images .
Update: I got it working with pure default Docker configuration (no manual DNS set to 8.8.8.8, no IP V6 deactivation), by removing Unlocator-related settings from the router (its specific DNS server, and maybe more problematic, static routes that redirected 8.8.8.8 and 8.8.4.4 to the router itself as recommended to get Chromecast working)! I had to reboot the router afterwards and disable/re-enable local network adapters. Strange that Docker is the only application which encountered issues because of this set-up.