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
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.
Fixed by following this advice:
[...] can you try to reset everything?
pkill docker
iptables -t nat -F
ifconfig docker0 down
brctl delbr docker0
docker -d
It will force docker to recreate the bridge and reinit all the network rules
https://github.com/dotcloud/docker/issues/866#issuecomment-19218300
Seems the interface was 'hung' somehow.
Update for more recent versions of docker:
The above answer might still get the job done for you but it has been quite a long time since this answer was posted and docker is more polished now so make sure you try these first before going into mangling with iptables and all.
sudo service docker restart or (if you are in a linux distro that does not use upstart) sudo systemctl restart docker
The intended way to restart docker is not to do it manually but use the service or systemctl command:
service docker restart
or
systemctl restart docker
Tried all answers, none worked for me.
After trying everything else I could find, this did the trick:
reboot
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.
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 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.
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.
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
Other answers have stated that the docker0 interface (bridge) can be the source of the problem. On Ubuntu 20.04 I observed that the interface was missing its IP address (to be checked with ip addr show dev docker0). Restarting Docker alone did not help. I had to delete the bridge interface manually.
sudo ip link delete docker0
sudo systemctl restart docker
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.
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
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
it help me:
sudo ip link delete docker0
sudo systemctl stop docker.socket
sudo systemctl stop docker.service
sudo systemctl start docker.socket
sudo systemctl start docker.service
NOTE: after this, interface docker0 must have ip adress smth like that:
inet 172.17.0.1/16 brd 172.17.255.255 scope global docker0
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"
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.
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
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
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
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.
On windows (8.1) I killed the virtualbox interface (via taskmgr) and it solved the issue.
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>
I've tried most answers in here, but the only thing that worked was re-creating the network:
$ docker network rm the-network
$ docker network create --driver=bridge the-network
I also needed to re-create the docker container that used it:
$ sudo docker create --name the-name --network the-network
Then it started with internet access.
I am on Arch Linux and after trying all the above answers I realized that I had a firewall enabled in my machine, nftables, and disabling it did the trick. I did :
sudo systemctl disable nftables
sudo systemctl stop nftables
sudo reboot
My network cards:
➜ ~ ip link
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN mode DEFAULT group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
2: enp1s0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc fq_codel state DOWN mode DEFAULT group default qlen 1000
link/ether 68:f7:28:84:e7:fe brd ff:ff:ff:ff:ff:ff
3: wlp2s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP mode DORMANT group default qlen 1000
link/ether d0:7e:35:d2:42:6d brd ff:ff:ff:ff:ff:ff
4: docker0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP mode DEFAULT group default
link/ether 02:42:43:3f:ff:94 brd ff:ff:ff:ff:ff:ff
5: br-c51881f83e32: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue state DOWN mode DEFAULT group default
link/ether 02:42:ae:34:49:c3 brd ff:ff:ff:ff:ff:ff
6: br-c5b2a1d25a86: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue state DOWN mode DEFAULT group default
link/ether 02:42:72:d3:6f:4d brd ff:ff:ff:ff:ff:ff
8: veth56f42a2@if7: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue master docker0 state UP mode DEFAULT group default
link/ether 8e:70:36:10:4e:83 brd ff:ff:ff:ff:ff:ff link-netnsid 0
and my firewal configuration, /etc/nftables.conf, which I now disabled and will futurely try to improve so I can have the docker0 network card rules setup correctly:
#!/usr/bin/nft -f
# vim:set ts=2 sw=2 et:
# IPv4/IPv6 Simple & Safe firewall ruleset.
# More examples in /usr/share/nftables/ and /usr/share/doc/nftables/examples/.
table inet filter
delete table inet filter
table inet filter {
chain input {
type filter hook input priority filter
policy drop
ct state invalid drop comment "early drop of invalid connections"
ct state {established, related} accept comment "allow tracked connections"
iifname lo accept comment "allow from loopback"
ip protocol icmp accept comment "allow icmp"
meta l4proto ipv6-icmp accept comment "allow icmp v6"
#tcp dport ssh accept comment "allow sshd"
pkttype host limit rate 5/second counter reject with icmpx type admin-prohibited
counter
}
chain forward {
type filter hook forward priority filter
policy drop
}
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.
There are lot of good answer already. I faced similar problem in my orange pi pc running armbian recently. Docker container was blocked to internet. This command solve the problem in my case. So I like to share it
docker run --security-opt seccomp=unconfined imageName
The only thing that works on my computer (which 8.8.8.8 didn't work)
$ netstat -ntpl | grep :53
tcp 0 0 10.194.128.1:53 0.0.0.0:* LISTEN -
tcp 0 0 127.0.0.53:53 0.0.0.0:* LISTEN -
tcp 0 0 192.168.122.1:53 0.0.0.0:* LISTEN -
tcp 0 0 127.0.2.1:53 0.0.0.0:* LISTEN -
sudo vim /etc/docker/daemon.json based on the information above on your own computer{
"dns": ["192.168.122.1","10.194.128.1"]
}
sudo ip link delete docker0
sudo systemctl restart docker
If you're running Docker rootless and facing this issue, during its installation, iptables may not be configured properly, mainly because of using --skip-iptables option when Docker complained about iptables:
[ERROR] Missing system requirements. Run the following commands to
[ERROR] install the requirements and run this tool again.
[ERROR] Alternatively iptables checks can be disabled with --skip-iptables .
########## BEGIN ##########
sudo sh -eux <<EOF
# Load ip_tables module
modprobe ip_tables
EOF
########## END ##########
Let's check if this is the issue: is ip_tables kernel module loaded?
sudo modprobe ip_tables
If there is no output, this answer may not help you (you could try anyway). Otherwise, the output's something like this:
modprobe: FATAL: Module ip_tables not found in directory /lib/modules/5.18.9-200.fc36.x86_64
Let's fix it!
First, uninstall Docker rootless (no need to stop the service via systemctl, the script handles that):
dockerd-rootless-setuptool.sh uninstall --skip-iptables
Ensure iptables package is installed, although it's shipped by default by major distributions.
Now, make ip_tables module visible to modprobe and install it (thanks to this):
sudo depmod
sudo modprobe ip_tables
Now, re-install Docker rootless:
dockerd-rootless-setuptool.sh install
If it doesn't bother about iptables, you're done and problem should be fixed. Don't forget to enable the service (i.e. systemctl enable --user --now docker)!
For me iwas facing the same issue as user of redhat/centos/fedora using podman
firewall-cmd --zone=public --add-masquerade
firewall-cmd --permanent --zone=public --add-masquerade
for more firewalld and podman (or docker) – no internet in the container and could not resolve host
As we know, the default bridge is built on a different docker0 interface and its IP is different from the main IP of the system.
To access the Internet, after reaching the gateway related to docker0, the packet must follow the path from the main gateway of the system, but this packet transfer is not possible because in the Arch operating system has a package that installed along with Docker that controls the connections, the name of this package is nftables.
To solve the problem I did this (You can have better settings for nftables):
1- First, I ran sudo vim /etc/nftables.conf and opened the configuration file
2- In the section below, I changed the chain forward and allow access for docker0 interface (thanks to @Lenar Hoyt).
chain forward {
type filter hook forward priority filter
policy drop
ct state { established, related } accept
ct state invalid drop
iif docker0 accept
}
3- In the last step, we execute these two commands
sudo systemctl restart nftables
sudo systemctl restart docker
I encountered this issue, it turned out to be the Docker + iptables conflict when the networking was set to host. The host could be using the newer nftables whereas Docker could still be using the iptables-legacy. This GitHub issue issues to explain the problem. Additionally, our setup was Docker-in-Docker where the container inside the container had no internet.
Run the following commands in your container to verify if this is the case:
ping google.com # Should fail since your container will have no internet
iptables -L
iptables-legacy -L
If the Docker rules are not present in iptables -L command output, then the conflict exists and we need to solve it.
Using update-alternatives to switch to iptables-legacy which would also be available in your distribution: https://wiki.debian.org/iptables#Current_status
Be sure to restart Docker after changing to iptables-legacy for this work,
update-alternatives --set iptables /usr/sbin/iptables-legacy
update-alternatives --set ip6tables /usr/sbin/ip6tables-legacy
If you are having this problem while building a Docker-in-Docker container then an easy workaround is to build using --network=host.
Bash solution to solve the conflict: https://github.com/garutilorenzo/iptables-docker if nothing works. Haven't tried, let me know if this worked for you.
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.
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
The issue was of linux using old version of libseccomp2
rpi ~$ sudo apt-key adv --keyserver keyserver.ubuntu.com --recv-keys 04EE7237B7D453EC 648ACFD622F3D138
rpi ~$ echo 'deb http://httpredir.debian.org/debian buster-backports main contrib non-free' | sudo tee -a /etc/apt/sources.list.d/debian-backports.list
rpi ~$ sudo apt update
rpi ~$ sudo apt install libseccomp2 -t buster-backports
After this try
rpi ~$ docker run -it --rm alpine:3.15.0
(alpine shell)# apk update
apk update will fetch and hence your are connected to internet
I was using
Linux raspberrypi 5.10.63-v7l+ #1496 SMP Wed Dec 1 15:58:56 GMT 2021 armv7l GNU/Linux
you may check using uname -a
In my case, for some unknown reason, docker was configured (by someone else) to not generate the iptables rules containers need for networking. The docker container could only ping the host, nothing else. Most answers here did not help; recreating the bridge nor restarting the service changed anything.
But then, by pure chance, I found the following:
$ cat /etc/docker/daemon.json
{"iptables": false}
I deleted the file /etc/docker/daemon.json and restarted the daemon (default is to create the iptables rules). This solved the networking problem.
Below, the iptables rules before the fix:
$ iptables -L -n
Chain INPUT (policy ACCEPT)
target prot opt source destination
Chain FORWARD (policy DROP)
target prot opt source destination
Chain OUTPUT (policy ACCEPT)
target prot opt source destination
Now, the iptables rules after the fix:
$ iptables -L -n
Chain INPUT (policy ACCEPT)
target prot opt source destination
Chain FORWARD (policy DROP)
target prot opt source destination
DOCKER-USER all -- 0.0.0.0/0 0.0.0.0/0
DOCKER-ISOLATION-STAGE-1 all -- 0.0.0.0/0 0.0.0.0/0
ACCEPT all -- 0.0.0.0/0 0.0.0.0/0 ctstate RELATED,ESTABLISHED
DOCKER all -- 0.0.0.0/0 0.0.0.0/0
ACCEPT all -- 0.0.0.0/0 0.0.0.0/0
ACCEPT all -- 0.0.0.0/0 0.0.0.0/0
ACCEPT all -- 0.0.0.0/0 0.0.0.0/0 ctstate RELATED,ESTABLISHED
DOCKER all -- 0.0.0.0/0 0.0.0.0/0
ACCEPT all -- 0.0.0.0/0 0.0.0.0/0
ACCEPT all -- 0.0.0.0/0 0.0.0.0/0
Chain OUTPUT (policy ACCEPT)
target prot opt source destination
Chain DOCKER (2 references)
target prot opt source destination
ACCEPT tcp -- 0.0.0.0/0 172.17.0.2 tcp dpt:8443
Chain DOCKER-ISOLATION-STAGE-1 (1 references)
target prot opt source destination
DOCKER-ISOLATION-STAGE-2 all -- 0.0.0.0/0 0.0.0.0/0
DOCKER-ISOLATION-STAGE-2 all -- 0.0.0.0/0 0.0.0.0/0
RETURN all -- 0.0.0.0/0 0.0.0.0/0
Chain DOCKER-ISOLATION-STAGE-2 (2 references)
target prot opt source destination
DROP all -- 0.0.0.0/0 0.0.0.0/0
DROP all -- 0.0.0.0/0 0.0.0.0/0
RETURN all -- 0.0.0.0/0 0.0.0.0/0
Chain DOCKER-USER (1 references)
target prot opt source destination
RETURN all -- 0.0.0.0/0 0.0.0.0/0
I just solve that problem, my container $container didn't have internet connection but it wasn't nothing to do with DNS because not even ping 8.8.8.8 was working. For a unknown reason the traffic of the container wasn't forwarding to my physical gateway so it wasn't out of my host/PC. So, I just add a masquerade rule for NAT:
network_name=$(docker inspect --format '{{ range $key, $value := .NetworkSettings.Networks }}{{ $key }}{{ end }}' $container)
etwork_id=$(docker network inspect -f {{.Id}} $network_name)
bridge_name="br-${network_id:0:12}"
IP=$(ip addr show $bridge_name | grep 'inet ' | awk '{print $2}')
sudo iptables -t nat -A POSTROUTING ! -o $bridge_name -s $IP -j MASQUERADE
user@host:~$ docker exec -u 0 -it $container /bin/bash
root@772bc683889a:/var/www/html# ping 8.8.8.8
PING 8.8.8.8 (8.8.8.8) 56(84) bytes of data.
64 bytes from 8.8.8.8: icmp_seq=1 ttl=118 time=12.1 ms
64 bytes from 8.8.8.8: icmp_seq=2 ttl=118 time=9.50 ms
64 bytes from 8.8.8.8: icmp_seq=3 ttl=118 time=4.98 ms
^C
--- 8.8.8.8 ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 2003ms
rtt min/avg/max/mdev = 4.979/8.850/12.070/2.931 m
