Docker registry error response from daemon i/o timeout

Question

Error response from daemon: Get https://registry.xxx.com:5000/v1/users/: dial tcp: lookup registry.xxx.com on 192.168.65.1:53: read udp 192.168.65.2:42307->192.168.65.1:53: i/o timeout

I have set up a private docker registry running on an Ubuntu server, but keep getting the error above when I do "docker login https://registry.xxx.com:5000".

When I run "curl -v https://registry.xxx.com:5000" from my mobile phone's network I can resolve the host and am able to login to the registry succesfully. But I cannot do this and get a "cannot resolve host" error when running "curl -v https://registry.xxx.com:5000 from the same network the registry is on".

It seems that it can't resolve the domain. I have a draytek router.

Any ideas?

Dread

Answer 1

Even though this problem was solved by a router change, I'm hit by the "read udp ... timeout" problem every other day. The bug is discussed in docker forums and is apparently related to a DNS lookup. I've found a work-around by doing a docker-machine restart.

Irresolute

Answer 2

Go to: Docker App -> Settings -> Resource -> Network -> DNS Server and set IP again.

I encounter this problem while I am at the company. no problem when at home.

I think it is the problem of the firewall in the company

Sequential

Idk why but I had a different value in "Docker subnet". I simply reset this to default value which is in my case "192.168.65.0/28" and it seems to work now. – Unthinkable

Answer 3

I've solved the issue by login with my docker id.

By default, my docker logged out. I go unnoticed. After logging into Docker, everything is running smoothly.

Bornu

Answer 4

Problem solved I had the registry.domain.com DNS-forwarded in the Draytek router, oops ;)

Dread

Answer 5

I've solved the exact same issue by force-ending the VBoxHeadless.exe process in the task manager.

Thorbert

Answer 6

I have solved the same issue by resetting the router dns cache, I suggest to try it.

Unanswerable