Docker Swarm - network sandbox join failed: subnet sandbox join failed for "10.0.0.0/24": error creating vxlan interface: file exists

Question

I met this issue when deploying my application on Docker Swarm.

I run docker stack deploy -c /path and get error network sandbox join failed: subnet sandbox join failed for "10.0.0.0/24": error creating vxlan interface: file exists

I searched too many pages in Google but not found any solution.

Weir

Answer 1

You can try to the following steps.

1. Check each node for any vx-* interfaces in /sys/class/net

$ ls -l /sys/class/net/ | grep vx

2. Once we have interface id's pull more details

$ udevadm info /sys/class/net/

3. If these interfaces exist we should be able to safely remove them. Replace vx-000000-xxxxx with the interface id from Step 2

$ sudo ip -d link show vx-000000-xxxxx
$ sudo ip link delete vx-000000-xxxxx

etc...

4. Redeploy the service.

Freed

I had the same issue and these steps solved it, thank you very much! – Dissociation

In my case i do not get usefull info in step 2. But you solved my problem. Thank you! – Fuller

Answer 2

If you are on a raspberry pi with Ubuntu you may just be missing the necessary libraries. You can install them with:

sudo apt install linux-modules-extra-raspi

Armand

worked for me (Ubuntu 22.04) ! ty – Insuperable

That was it for me, too. (Ubuntu 22.04 LTS on a Pi4) – Hughs

Answer 3

There might be conflict between multiple networks on your host

Did you check how many dockerd services is running? There might be two docker service is running on your host, via apt and snap

To remove docker via snap:

sudo snap remove docker

or, to avoid it creating back-up data:

sudo snap remove --purge docker

To remove docker via apt

sudo apt-get remove docker docker-engine docker.io containerd runc

And then reinstalled from here https://docs.docker.com/engine/install/ubuntu/

Pittsburgh

Answer 4

I'm running docker swarm, and occasionally (by every few weeks) a service fails to start giving out this error.

The fix for me is to restart docker on the docker swarm node where the error is occuring. This seems to "free up" the network subnet, which causes the service/container to start successfully on another node automatically. So, in other words, please run

sudo systemctl restart docker  

Senile