Red Hat Bugzilla – Bug 1249701
Images fail to download from time to time, restarting docker.service helps
Last modified: 2016-02-12 07:59:07 EST
Description of problem:
Sometimes, downloading a docker image fails. This wouldn't be that much of a problem, if it could be quickly restarted, but the timeout seems to be over a minute and Ctrl+C doesn't kill the attempt to download (runs in the background) and prevents a retry, until the last one finishes.
Version-Release number of selected component (if applicable):
Docker version 1.7.1.fc22, build b6416b7/1.7.1
Steps to Reproduce:
1. Make sure not to have the containers for the image downloaded.
2. docker run -it fedora:latest bash
Unable to find image 'fedora:latest' locally
Trying to pull repository docker.io/fedora ... failed
Get https://index.docker.io/v1/repositories/library/fedora/images: dial tcp: lookup index.docker.io: No address associated with hostname
Successful download and run.
The only way to retry quickly, is to restart docker.service. After this, it usually works, so I am sure it's not a docker.io blackout.
I have no idea what could cause this.
Michal any ideas?
Are you still seeing this problem with docker-1.8.2?
Viktor is no longer in RH, so clearing the needinfo.