I had it working allright but now it stopped. I tried the following commands with no avail:
docker run -dns 8.8.8.8 base ping google.com
docker run base ping google.com
sysctl -w net.ipv4.ip_forward=1
- both on the host and on the container
All I get is unknown host google.com
. Docker version 0.7.0
Any ideas?
P.S. ufw
disabled as well
This question is related to
docker
No internet access can also be caused by missing proxy settings. In that case, --network host
may not work either. The proxy can be configured by setting the environment variables http_proxy
and https_proxy
:
docker run -e "http_proxy=YOUR-PROXY" \
-e "https_proxy=YOUR-PROXY"\
-e "no_proxy=localhost,127.0.0.1" ...
Do not forget to set no_proxy as well, or all requests (including those to localhost) will go through the proxy.
More information: Proxy Settings in the Archlinux Wiki.
If you're on OSX, you might need to restart your machine after installing Docker. This has been an issue at times.
For me it was an iptables forwarding rule. For some reason the following rule, when coupled with docker's iptables rules, caused all outbound traffic from containers to hit localhost:8080
:
iptables -t nat -A PREROUTING -p tcp --dport 80 -j REDIRECT --to-ports 8080
iptables -t nat -I OUTPUT -p tcp -d 127.0.0.1 --dport 80 -j REDIRECT --to-ports 8080
for me, my problem was because of iptables-services was not installed, this worked for me (CentOS):
sudo yum install iptables-services
sudo service docker restart
I do not know what I am doing but that worked for me :
OTHER_BRIDGE=br-xxxxx # this is the other random docker bridge (`ip addr` to find)
service docker stop
ip link set dev $OTHER_BRIDGE down
ip link set dev docker0 down
ip link delete $OTHER_BRIDGE type bridge
ip link delete docker0 type bridge
service docker start && service docker stop
iptables -t nat -A POSTROUTING ! -o docker0 -s 172.17.0.0/16 -j MASQUERADE
iptables -t nat -A POSTROUTING ! -o docker0 -s 172.18.0.0/16 -j MASQUERADE
service docker start
I've had a similar problem for the last few days. For me the cause was a combination of systemd, docker and my hosting provider. I'm running up-to-date CentOS (7.7.1908).
My hosting provider automatically generates a config file for systemd-networkd. Starting with systemd 219 which is the current version for CentOS 7, systemd-networkd took control of network-related sysctl parameters. Docker seems to be incompatible with this version and will reset the IP-Forwarding flags everytime a container is launched.
My solution was to add IPForward=true
in the [Network]
-section of my provider-generated config file. This file might be in several places, most likely in /etc/systemd/network
.
The process is also described in the official docker docs: https://docs.docker.com/v17.09/engine/installation/linux/linux-postinstall/#ip-forwarding-problems
For Ubuntu 19.04 using openconnect 8.3 for VPN, I had to symlink /etc/resolve.conf to the one in systemd (opposite of answerby wisbucky )
sudo ln -sf /etc/resolv.conf /run/systemd/resolve/resolv.conf
Steps to debug
Docker version: Docker version 19.03.0-rc2, build f97efcc
Originally my docker container was able to reach the external internet (This is a docker service/container running on an Amazon EC2).
Since my app is an API, I followed up the creation of my container (it succeeded in pulling all the packages it needed) with updating my IP Tables to route all traffic from port 80 to the port that my API (running on docker) was listening on.
Then, later when I tried rebuilding the container it failed. After much struggle, I discovered that my previous step (setting the IPTable port forwarding rule) messed up the docker's external networking capability.
Solution: Stop your IPTable service:
sudo service iptables stop
Restart The Docker Daemon:
sudo service docker restart
Then, try rebuilding your container. Hope this helps.
Follow Up
I completely overlooked that I did not need to mess with the IP Tables to forward incoming traffic to 80 to the port that the API running on docker was running on. Instead, I just aliased port 80 to the port the API in docker was running on:
docker run -d -p 80:<api_port> <image>:<tag> <command to start api>
Sharing a simple and working solution for posterity. When we run a docker container without explicitly mentioning the --network
flag, it connects to its default bridge network which prohibits connecting to the outside world. To resolve this issue, we have to create our own bridge network(user-defined bridge) and have to explicitly mention it with the docker run command.
docker network create --driver bridge mynetwork
docker run -it --network mynetwork image:version
for me, using centos 7.4, it was not issue of /etc/resolve.conf, iptables, iptables nat rules nor docker itself. The issue is the host missing the package bridge-utils which docker require to build the bridge using command brctl. yum install -y bridge-utils and restart docker, solve the problem.
First thing to check is run cat /etc/resolv.conf
in the docker container. If it has an invalid DNS server, such as nameserver 127.0.x.x
, then the container will not be able to resolve the domain names into ip addresses, so ping google.com
will fail.
Second thing to check is run cat /etc/resolv.conf
on the host machine. Docker basically copies the host's /etc/resolv.conf
to the container everytime a container is started. So if the host's /etc/resolv.conf
is wrong, then so will the docker container.
If you have found that the host's /etc/resolv.conf
is wrong, then you have 2 options:
Hardcode the DNS server in daemon.json. This is easy, but not ideal if you expect the DNS server to change.
Fix the hosts's /etc/resolv.conf
. This is a little trickier, but it is generated dynamically, and you are not hardcoding the DNS server.
1. Hardcode DNS server in docker daemon.json
Edit /etc/docker/daemon.json
{
"dns": ["10.1.2.3", "8.8.8.8"]
}
Restart the docker daemon for those changes to take effect:
sudo systemctl restart docker
Now when you run/start a container, docker will populate /etc/resolv.conf
with the values from daemon.json
.
2. Fix the hosts's /etc/resolv.conf
A. Ubuntu 16.04 and earlier
For Ubuntu 16.04 and earlier, /etc/resolv.conf
was dynamically generated by NetworkManager.
Comment out the line dns=dnsmasq
(with a #
) in /etc/NetworkManager/NetworkManager.conf
Restart the NetworkManager to regenerate /etc/resolv.conf
:
sudo systemctl restart network-manager
Verify on the host: cat /etc/resolv.conf
B. Ubuntu 18.04 and later
Ubuntu 18.04 changed to use systemd-resolved
to generate /etc/resolv.conf
. Now by default it uses a local DNS cache 127.0.0.53. That will not work inside a container, so Docker will default to Google's 8.8.8.8 DNS server, which may break for people behind a firewall.
/etc/resolv.conf
is actually a symlink (ls -l /etc/resolv.conf
) which points to /run/systemd/resolve/stub-resolv.conf
(127.0.0.53) by default in Ubuntu 18.04.
Just change the symlink to point to /run/systemd/resolve/resolv.conf
, which lists the real DNS servers:
sudo ln -sf /run/systemd/resolve/resolv.conf /etc/resolv.conf
Verify on the host: cat /etc/resolv.conf
Now you should have a valid /etc/resolv.conf
on the host for docker to copy into the containers.
You may have started your docker with dns options --dns 172.x.x.x
I had the same error and removed the options from /etc/default/docker
The lines:
# Use DOCKER_OPTS to modify the daemon startup options.
DOCKER_OPTS="--dns 172.x.x.x"
I also encountered such an issue while trying to set up a project using Docker-Compose on Ubuntu.
The Docker had no access to internet at all, when I tried to ping any IP address or nslookup some URL - it failed all the time.
I tried all the possible solutions with DNS resolution described above to no avail.
I spent the whole day trying to find out what the heck is going on, and finally found out that the cause of all the trouble was the antivirus, in particular it's firewall which for some reason blocked Docker from getting the IP address and port.
When I disabled it - everything worked fine.
So, if you have an antivirus installed and nothing helps fix the issue - the problem could be the firewall of the antivirus.
On centos 8, My problem was that I did not install & start iptables before starting docker service. Make sure iptables service is up and running before you start docker service.
The intended way to restart docker is not to do it manually but use the service
or init command:
service docker restart
I was using DOCKER_OPTS="--dns 8.8.8.8"
and later discovered and that my container didn't have direct access to internet but could access my corporate intranet. I changed DOCKER_OPTS
to the following:
DOCKER_OPTS="--dns <internal_corporate_dns_address"
replacing internal_corporate_dns_address
with the IP address or FQDN of our DNS and restarted docker using
sudo service docker restart
and then spawned my container and checked that it had access to internet.
I had the problem on Ubuntu 18.04. However the problem was with the DNS. I was in a corporate network that has its own DNS server and block other DNS servers. This is to block some websites (porn, torrents, ... so on )
To resolve your problem
use --dns your_dns as suggested by @jobin
docker run --dns your_dns -it --name cowsay --hostname cowsay debian bash
Just adding this here in case someone runs into this issue within a virtualbox container running docker. I reconfigured the virtualbox network to bridged instead of nat, and the problem went away.
I was stumped when this happened randomly for me for one of my containers, while the other containers were fine. The container was attached to at least one non-internal network, so there was nothing wrong with the Compose
definition. Restarting the VM / docker daemon did not help. It was also not a DNS issue because the container could not even ping
an external IP. What solved it for me was to recreate the docker network(s). In my case, docker-compose down && docker-compose up
worked.
This forces the recreation of all networks of all the containers:
docker-compose down
&& docker-compose up
I suppose you just remove and recreate the service, which recreates the service's network(s):
docker service rm some-service
docker service create ...
Simply remove and recreate the external networks of that service:
docker network rm some-external-network
docker network create some-external-network
For me it was the host's firewall. I had to allow DNS on the host's firewall. And also had to restart docker after changing the host firewall setting.
Updating this question with an answer for OSX (using Docker Machine)
If you are running Docker on OSX using Docker Machine, then the following worked for me:
docker-machine restart
<...wait for it to restart, which takes up to a minute...>
docker-machine env
eval $(docker-machine env)
Then (at least in my experience), if you ping google.com from a container all will be well.
On windows (8.1) I killed the virtualbox interface (via taskmgr) and it solved the issue.
Source: Stackoverflow.com