-
Notifications
You must be signed in to change notification settings - Fork 18.6k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Docker fails to create network bridge on start up #18113
Comments
@Chili-Man this seems like a dupe of #17939. Can you pls make sure if there are multiple docker daemons running (either natively in the host or via dind using the same root directory as indicated in #17939). |
@mavenugo At first I thought that the issue was that there were multiple docker daemons running, but that is not that case. I've checked all the processes running on the host machine, and there was no docker daemon running at all. |
@Chili-Man do you know why the log indicates the daemon is launched multiple times within the same second ( |
yeah its within the same second because docker fails fast and upstart attempts a couple of times to restart it after it fails to start; I also manually tried starting the docker daemon after upstart gave up on trying with:
but it still produced the same error. I made sure that no other docker daemon was running when I issued the above command. It is strange that I haven’t been able to reproduce that issue with the 3.19 kernel, only on the 4.2.0 kernel. |
@Chili-Man are you still having this issue? 1.9.1 was released recently; wondering if that release solved the issue for you, or if it's still unresolved. |
I haven't had a chance to try it out with 1.9.1, I'll try it over this weekend and update back here when I do. |
@Chili-Man thanks! Not sure if there's anything changed in 1.9.1 for this particular case, but there were quite some changes, so interested to hear if it's still an issue |
I'm having the same issue with 1.9.1. It's intermittent across my organization. I've put some information below along the lines of Chili-Man's supplied information.
|
I am also experiencing this issue Client: INFO[0000] API listen on /var/run/docker.sock |
ping @aboch |
Most likely something is messing the @mavenugo moby/libnetwork#687 would help give some clarity on where the failure is, as I have said in #17939 (comment) |
@Chili-Man @dvanbuskirk @sheldonkwok A possible work-around, please check, could be delete that file before starting the daemon. |
Deleting the file worked... Thanks! |
@sheldonkwok Thanks for confirming. |
I have it on s3. There's nothing sensitive in it right? I glanced through it and it looked fine |
Yeah, nothing sensitive in it. Mostly Network, Endpoint and Sandbox ids and their configurations. Thanks. |
Sure, me@sheldonkwok.com |
Thanks @sheldonkwok for sending the file, it was very helpful. In your case the issue was because the local store file had some inconsistent states, due to a sequence of ungraceful daemon shutdown. moby/libnetwork#794 was pushed to restrict the chances of ending in such inconsistent file state. Next libnetwork vendoring will bring it to docker/docker. |
Great thanks for the quick response and the temporary solution on my end |
I have this problem intermittently as well. |
Now that moby/libnetwork#687 is merged on master, at least for some cases, the no This means the local-kv.db file got somehow overwritten probably by another daemon instance. A way to say you are hitting this issue is to check the o/p of |
@aboch, I just hit this issue when upgrading our 1.9.0 machines to 1.9.1 on Ubuntu 14.04. "Fix" was to delete the:
If you guys need another local-kv.db file to look at, I can send it to you. |
Had same problem with ubuntu 15.04 and docker version 1.9.1 deleting local-kv.db fixed it |
Deleting |
I was experiencing the same problem. Removing |
Removing /var/lib/docker/network/files/local-kv.db as root did not work for me for those errors:
ERROR: for gateway-app Cannot start service gateway-app: failed to update store for object type *libnetwork.endpointCnt: Key not found in store ERROR: for onboarding-app Cannot start service onboarding-app: failed to update store for object type *libnetwork.endpointCnt: Key not found in store ERROR: for hb-rabbitmq Cannot start service hb-rabbitmq: service endpoint with name dev_hb-rabbitmq_1 already exists ERROR: for hb-mongodb Cannot start service hb-mongodb: failed to update store for object type *libnetwork.endpointCnt: Key not found in |
Just ran into the same issue with docker 1.11.2 sudo systemctl status docker gave me:
Deleting /var/lib/docker/network/files/local-kv.db fixed the problem and allowed me to start up docker successfully. This is quite a serious issue, it's taking down production machines on our Google Container Engine. Detailed Version Info:
|
Had this problem too; Debian Jessie and 1.11.2, exactly as described by the previous comment. Removing |
Encountered same problem with docker 1.9.1 and CentOS 7 with Linux kernal 3.18.27. deleting /var/lib/docker/network/files/local-kv.db works for me.
|
+1, deleting |
Had this problem too; Ubuntu 14.04 and 1.12.6, exactly as described by the previous comment. Removing /var/lib/docker/network/files/local-kv.db allowed me to start docker containers again. |
Same with Ubuntu 16.04.1 and Docker 1.11.2 -> Deleting local-kv.db safed the day... |
This fixed it for me:
|
docker 1.13.0 cannot start if I have an IPv6 nameserver in /etc/resolv.conf !?! |
All the solution tried, still face this error: |
Yes, I found programster's conclusion is right. If the VPN is working, the Docker installation process will fail. After disabling the VPN, it works. So it means when the VPN is working, Docker's installation logic couldn't correctly handle the network status. |
Remove all ipv6 nameservers from /etc/resolv.conf!
clockzhong <notifications@github.com> ezt írta (időpont: 2017. márc. 3., P
5:43):
… All the solution tried, still face this error:
Error starting daemon: Error initializing network controller: list bridge
addresses failed: no available network
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#18113 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AAPoSvKCeIncEpGKSLQbrTx58I62YOoOks5rh5pigaJpZM4GmJnl>
.
|
What I've found is that if there's a n ipv6 nameserver in /etc/resolv.conf,
then docker won't start.
This can be a result of a started VPN, though.
clockzhong <notifications@github.com> ezt írta (időpont: 2017. márc. 3., P
5:46):
… Yes, I found programster's conclusion is right. If the VPN is working, the
Docker installation process will fail. After disabling the VPN, it works.
So it means when the VPN is working, Docker's installation logic couldn't
correctly handle the network status.
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#18113 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AAPoSoJwxPVPCW44CMjy5E4tIVqdVG9rks5rh5sxgaJpZM4GmJnl>
.
|
Tgulacsi, my failure doesn't relate to the /etc/resolv.conf. Even this file doesn't have the IPV6 nameserver in my envriroment, it still fails when my VPN is enabling. It seems this is a multiple-root-causes bug in Docker's installation process. |
@tgulacsi If the nameserver contains a zone id (i.e., the |
I experienced the same problem with Docker 17.06, Ubuntu 16 (Linux kernel 4.4.0), on a Docker swarm worker node. Removing
This is happening with a worker node. The container fails to start with this inspect log:
And I got a bunch of error logs like this:
@Chili-Man Would you mind re-opening this issue please? Thank you. |
@osman-masood please open a new issue with details; the original issue is from 2015, at which time the Swarm-managed overlay networks didn't even exist, so your issue may well be unrelated (but similar outcome) |
On MacOS, Restarting the docker daemon did it for me. |
will fix it |
As there are several tickets written by people claming that IPv6 is broken, here is the solution how to configure it corretly. Someone should change the documentation accordingly. Docker IPv6 is working perfectly, it's just the documentation that doesn't tell people what to do in order to set it up correctly. Docker will not do SLAAC (Stateless Address Autoconfiguration), nor will it use DHCPv6. As it also won't do NATv6 (thank god!), you will have to assign ipv6 prefixes to your networks manually. The default bridge network has to be assigned a network, you can do this by adding the following to daemon.json:
For any other network you might have, you can decide wether or not it will be ipv6 capable. To create an IPv6 capable network, do the following: Also remember, that this is not NATv6 (thank god!). Therefor you will have to have a route from your router to your docker host! |
@hr1232 if you know which page(s) of the documentation contain the incorrect information, could you perhaps open a ticket in the documentation repository (or, even better; if you're interested in contributing, open a pull request 😅). If the incorrect information is in the daemon reference page, or one of the CLI reference pages, the markdown file for that file can be found in the cli repository; https://github.com/docker/cli/blob/master/docs/reference/commandline/dockerd.md, other pages are built from the documentation repository; https://github.com/docker/docker.github.io/ |
Description of problem:
When freshly installing the docker 1.9.0 daemon, it sometimes fails to create the network bridge at startup and thus fails to start the daemon. It seems to fail about 50% of the time and I'm not sure why. Here's some of the log output:
docker version
:docker info
:Cannot connect to the Docker daemon. Is the docker daemon running on this host?
uname -a
:Linux default-ubuntu-1404 4.2.0-18-generic #22~14.04.1-Ubuntu SMP Fri Nov 6 22:20:11 UTC 2015 x86_64 x86_64 x86_64 GNU/Linux
Environment details (AWS, VirtualBox, physical, etc.):
I've ran into this issue on both AWS and Virtualbox with the latest ubuntu 14.04 images with the
linux-virtual-lts-wily
kernels installed (4.2.0)How reproducible:
It seems to be about a 50% chance that the install will fail due to not being able to create the network bridge interface at start up
Steps to Reproduce:
linux-virtual-lts-wily
package onto an ubuntu 14.04 virtual box or AWS server-s overlay
Actual Results:
Docker does not run
Expected Results:
Docker daemon should be running
Additional info:
I dont get this error if I use the
linux-virtual-lts-vivid
kernel (3.19) instead.The text was updated successfully, but these errors were encountered: