Open Coolfeather2 opened 3 years ago
Same issue. For me, this only happens if the file /etc/docker/daemon.json exists with the following content:
{
"userland-proxy": false,
"ipv6": true,
"fixed-cidr-v6": "fd00::/80"
}
daemon.json contains:
{
"storage-driver": "overlay2",
"userland-proxy": false,
"log-driver": "json-file",
"init": true,
"log-opts": {"max-size": "10m", "max-file": "3"}
}
I suspect this would be a configuration issue for the LXC container; have you asked in the LXC forum? https://discuss.linuxcontainers.org/c/lxc/6
I'm able to run docker without applying the daemon, something in the daemon is messing with docker and removing the daemon after doesn't fix the issue
Isolated the cause to "userland-proxy": false
in the daemon.json
Expected behavior
Start Docker service
Actual behavior
Docker service fails to start
(code=exited, status=1/FAILURE)
Steps to reproduce the behavior
Installed docker on new Ubuntu 18.04 LXC Container Docker service would not start, reporting the error to journalctl Tested by running just
dockerd
in console and same error occuredOutput of
docker version
:Output of
docker info
:Docker Log:
Additional environment details (AWS, VirtualBox, physical, etc.) Much like #1033, I am having a similar issue but this also has /proc/sys/net/ipv4/conf/docker0/route_localnet being read-only Initally the host rebooted and docker wouldn't start, I created a new LXC Container to test if it was an issue with the container but the same issue occured in the new one.
System information:
Running inside LXC on Proxmox Fuse, KeyCTL and nesting enabled.