libnetwork: Cannot start container: subnet sandbox join failed for "10.0.0.0/24": error creating vxlan interface: file exists

(Similar, but closed, issues: #562, #751)

Very occasionally, I see this error message when starting a container in my swarm:

Error response from daemon: Cannot start container <container hash>: subnet sandbox join failed for "10.0.0.0/24": error creating vxlan interface: file exists

This error persists until I reboot the docker hosts. A comment on #751 suggested that restarting iptables would suffice; I have not tried this yet. I also have tried the solution mentioned in #562 previously, and I believe that worked as well, but I cannot remember for sure.

docker version:

Client:
 Version:      1.10.0
 API version:  1.22
 Go version:   go1.5.3
 Git commit:   590d5108
 Built:        Thu Feb  4 18:18:11 2016
 OS/Arch:      darwin/amd64

Server:
 Version:      swarm/1.1.0
 API version:  1.22
 Go version:   go1.5.3
 Git commit:   a0fd82b
 Built:        Thu Feb  4 08:55:18 UTC 2016
 OS/Arch:      linux/amd64

docker info:

Note: this was not captured when the error was occurring. If it happens again, I will comment with the info.

Containers: 32
 Running: 29
 Paused: 0
 Stopped: 3
Images: 222
Role: primary
Strategy: spread
Filters: health, port, dependency, affinity, constraint
Nodes: 4
 druid-vms.historical1: <host3>:2376
  └ Status: Healthy
  └ Containers: 4
  └ Reserved CPUs: 0 / 2
  └ Reserved Memory: 0 B / 8.188 GiB
  └ Labels: executiondriver=native-0.2, historical=yes, kernelversion=3.16.0-30-generic, operatingsystem=Ubuntu 14.04.2 LTS, provider=generic, storagedriver=aufs
  └ Error: (none)
  └ UpdatedAt: 2016-02-11T15:00:03Z
 druid-vms.historical2: <host4>:2376
  └ Status: Healthy
  └ Containers: 4
  └ Reserved CPUs: 0 / 2
  └ Reserved Memory: 0 B / 8.188 GiB
  └ Labels: executiondriver=native-0.2, historical=yes, kernelversion=3.16.0-30-generic, operatingsystem=Ubuntu 14.04.2 LTS, provider=generic, storagedriver=aufs
  └ Error: (none)
  └ UpdatedAt: 2016-02-11T15:00:24Z
 druid-vms.ingestion1: <host2>:2376
  └ Status: Healthy
  └ Containers: 5
  └ Reserved CPUs: 0 / 2
  └ Reserved Memory: 0 B / 8.188 GiB
  └ Labels: executiondriver=native-0.2, ingestion=yes, kernelversion=3.16.0-30-generic, operatingsystem=Ubuntu 14.04.2 LTS, provider=generic, storagedriver=aufs
  └ Error: (none)
  └ UpdatedAt: 2016-02-11T14:59:45Z
 druid-vms.queen: <host1>:2376
  └ Status: Healthy
  └ Containers: 19
  └ Reserved CPUs: 0 / 2
  └ Reserved Memory: 0 B / 8.188 GiB
  └ Labels: broker=yes, consul=yes, deployment-master=yes, executiondriver=native-0.2, kernelversion=3.16.0-30-generic, metadata=yes, operatingsystem=Ubuntu 14.04.2 LTS, provider=generic, storagedriver=aufs, swarm-master=yes
  └ Error: (none)
  └ UpdatedAt: 2016-02-11T15:00:22Z
Plugins: 
 Volume: 
 Network: 
Kernel Version: 3.16.0-30-generic
Operating System: linux
Architecture: amd64
CPUs: 8
Total Memory: 32.75 GiB
Name: druid-vms.queen
Http Proxy: <snip>
Https Proxy: <snip>
No Proxy: <snip>

uname -a:

• Local: Darwin <hostname> 15.3.0 Darwin Kernel Version 15.3.0: Thu Dec 10 18:40:58 PST 2015; root:xnu-3248.30.4~1/RELEASE_X86_64 x86_64
• Docker hosts: Linux <hostname> 3.16.0-30-generic #40~14.04.1-Ubuntu SMP Thu Jan 15 17:43:14 UTC 2015 x86_64 x86_64 x86_64 GNU/Linux

Environment details (AWS, VirtualBox, physical, etc.):

• Docker Swarm, 4 pre-existing VMs, generic driver

How reproducible:

• Intermittent; not reliably reproducible (and I wouldn’t begin to know how to try). Seems to occur sometime after destroying and re-creating the swarm using docker-machine.