problem while aborting a docker pull cmd

[garfee]

I restart the a pull cmd after abort it by ctrl^c, a error msg comes like this:

2013/12/08 12:30:20 Error: pull sameersbn/gitlab: is already in progress

I have to restart docker server to exec docker pull again.
I think it will be graceful to automatically detect and solve in docker itself.

[ratbeard]

I encountered this as well. The workaround of restarting the server process worked. The error message was a little different for me, probably why I didn't find the workaround sooner:

docker pull ubuntu
Repository ubuntu already being pulled by another client. Waiting.

A short term solution could be include the workaround in the error message, something like:

docker pull ubuntu
Repository ubuntu already being pulled by another client. Waiting.
If the previous pull of repository ubuntu was interrupted, restart the docker server and rerun the pull command.

[ratbeard]

Related #3136. It sounds like the download continues on in the background.

[bastengao]

Restart OS is work for me.

+1 -- "I think it will be graceful to automatically detect and solve in docker itself." -- yes.

[sfitts]

Just ran into this myself via the REST API. For use in a production environment I'd like to see at least some way to get the list of pending pulls from the server and the ability to abort any of them. That way at least you can manage the issue.

[jin]

+1. It'll be great to not have to restart the server whenever the pull fails halfway.

$ docker pull ubuntu
Pulling repository ubuntu
3db9c44f4520: Pulling dependent layers
0232356d9a18: Pulling dependent layers
9d076f5398dd: Pulling dependent layers
1edb91fcb5b5: Pulling dependent layers
2e68994ad893: Pulling dependent layers
c1bd2f6276a1: Pulling dependent layers
511136ea3c5a: Download complete
224eee0e1b55: Downloading [=======================>          ] 32.23 MB/67.48 MB 4m22s
2014/04/23 12:56:09 unexpected EOF

$ docker pull ubuntu
Repository ubuntu already being pulled by another client. Waiting.

[disposable-ksa98]

service docker restart worked for me.

[sfitts]

For me the issue isn't that you can't unwedge the server (clearly you can) it's that all its takes is one stray pull to be interrupted and you've just hung the server out to dry. That might be OK for the individual developer using their own docker server. But it really isn't what you want of a shared resource in production (IMO).

Exactly
On May 6, 2014 12:08 AM, "Sean Fitts" notifications@github.com wrote:

For me the issue isn't that you can't unwedge the server (clearly you can)
it's that all its takes is one stray pull to be interrupted and you've just
hung the server out to dry. That might be OK for the individual developer
using their own docker server. But it really isn't what you want of a
shared resource in production (IMO).

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/3115#issuecomment-42266512
.

[iambowen]

@sfitts agree! I met the same issue locally when I try to pull images. ps does not indicate which one of them is doing the pulling.

[nowthatsamatt]

So I had the same issue but after restarting I was able to simultaneously download the same image on two different boxes. So restarting gives you the ability for simultaneous downloads? :p

[j0hnsmith]

+1 restarting docker shouldn't be necessary after stopping a pull with ctrl+c

[majkinetor]

+1

[pantaoran]

+1 please fix this in Docker!