Docker CE 18.09.0 on Ubuntu 18.04.1 doesn't create any containers #485
Comments
|
@ata-sql could you provide more information?
I see that you have only 2 images, and no containers on that machine; Did you have existing containers before doing the dist-upgrade? If so; did you remove those after upgrading? Or did you perhaps switch to a different storage driver? |
|
a) 18.06.1, without any problems on ~100 PCs |
|
It only occurs on hosts that were upgraded, or do you see the same on a completely fresh host as well? |
|
What do you mean by that? We didn't upgrade OS distribution from 16.04 to 18.04 - it was 18.04 for some time (month or two) - and it wasn't upgrade of running system, but fresh reinstall of 18.04. |
|
Apologies, I think I misunderstood your mention of "after dist-upgrade of Ubuntu to Ubuntu 18.04.1 LTS", so I thought you updated an existing machine to 18.04.1 and docker stopped working. So; IIUC;
I'm not very familiar with Saltstack (and of course don't know how its configured in your situation);
I just tried an install on a fresh Ubuntu 18.04.1 machine, and was not able to reproduce the issue, so anything you can provide to narrow down what's causing the problem for you would be helpful. Here's the output of the machine I just installed ; |
|
daemon.json: check-config.sh output: /var/log/systlog: |
Not sure how this would relate to starting new containers, but I overlooked you're using live-restore; did you stop all containers before upgrading the docker daemon? live-restore cannot be used when doing major version upgrades (e.g. 18.06 -> 18.09), and can only be used for installing patch releases (e.g. I don't see anything standing out in the output of check-config.sh (all looks normal) I do see some messages about |
|
For sure there were no running containers when I did it in test environment. net=none |
|
No clue yet. Did a quick search for these errors; moby/moby#36084 mentions a similar error, and looks to be related to a specific kernel version. Other possible suspects would be the |
|
Adding and still |
|
It's not the |
|
Having the same issue on Debian 9.6 after upgrading from 18.03 or 18.06 to 18.09 ( |
|
Same issue on 18.04 for me. OCI runtime errors with latest docker package. Needed to downgrade to 18.06.1 The latest docker pkg is definitely borked somehow. |
|
@pisymbol did you get any error logs during startup of the dockerd (and/or containerd) service? |
|
Nothing visible. I will have to check journalctl. But the docker dpkg one
back works fine which leads me to packaging.
…On Tue, Nov 20, 2018 at 4:05 PM Sebastiaan van Stijn < ***@***.***> wrote:
@pisymbol <https://github.com/pisymbol> did you get any error logs during
startup of the dockerd (and/or containerd) service?
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#485 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/ARGxVuUQEHrixKmkmx0oQelw-zUZcPEKks5uxG6GgaJpZM4YbBni>
.
|
|
There was an issue with packaging due to the The |
|
I'm having this same problem since upgrading my ubuntu 16.04 server from 18.06 to 18.09. |
|
Fixed it by removing old version of runc - using whereis runc. I had the same issue after migration to 18.04.1 and docker-ce version is docker version The error is |
|
Is there any update to this? I don't have an old version of runc (I do have a custom script though called runc that calls docker run etc.). |
|
Are you planning to fix that issue? It's more than 2 months now and issue seems to be quite serious. |
|
@ata-sql do you happen to have a customised systemd unit file for the docker service? I just recalled docker/docs#8001 (related to moby/moby#38506), which could be related. Fixing the issue will be quite difficult if it's there's no consistent reproducer, or cause identified. |
|
So for me my fix was simply to rename my 'runc' script to something else and that fixed it. |
|
Hello, |
|
For others: |
|
Good to hear @ata-sql! I'll close this issue |
@ata-sql commented on Fri Nov 09 2018
hello,
after dist-upgrade of Ubuntu to
Ubuntu 18.04.1 LTSwe can't create any containersdocker run --network none --rm hello-worldends with:docker: Error response from daemon: OCI runtime create failed: container_linux.go:348: starting container process caused "exec: \"/hello\": stat /hello: no such file or directory": unknown.In syslog we have then:
Downgrading docker to
18.06.1-ce(only that package, rest of system remains unchanged) fixes this problem.The text was updated successfully, but these errors were encountered: