It will only try IPv6 if your computer has an address. Now the question is: if your network doesn't support it, then why do you have an address?
this post was submitted on 11 Aug 2025
46 points (96.0% liked)
Selfhosted
50502 readers
443 users here now
A place to share alternatives to popular online services that can be self-hosted without giving up privacy or locking you into a service you don't control.
Rules:
-
Be civil: we're here to support and learn from one another. Insults won't be tolerated. Flame wars are frowned upon.
-
No spam posting.
-
Posts have to be centered around self-hosting. There are other communities for discussing hardware or home computing. If it's not obvious why your post topic revolves around selfhosting, please include details to make it clear.
-
Don't duplicate the full text of your blog or github here. Just post the link for folks to click.
-
Submission headline should match the article title (don’t cherry-pick information from the title to fit your agenda).
-
No trolling.
Resources:
- selfh.st Newsletter and index of selfhosted software and apps
- awesome-selfhosted software
- awesome-sysadmin resources
- Self-Hosted Podcast from Jupiter Broadcasting
Any issues on the community? Report it using the report flag.
Questions? DM the mods!
founded 2 years ago
MODERATORS
They said thier internet connection didn't support it, not their network.
You can lease ipv6 all day long but its not going to go anywhere if it can't route to the internet.
Well where is the router getting the prefix from if their ISP doesn't support it.
Some routers advertise a routable link local.
Link local adresses.
Docker is a distraction in your problem description.
It's like if you asked why the top gear in your car isn't working and gave the model of car and engine type and gearbox. But it's really that you're stuck in slow traffic. Focus on the road name and destination to find a faster route.
For your problem, search for how to disable ipv6 for the Linux distribution and version that you have installed. You will find lots of guidance. Or share those details here for someone to help.
Or, better might be to see if there is a way to get ipv6 tunneling working on your connection. It may be possible even if the ISP is unhelpful.
It’s like if you asked why the top gear in your car isn’t working and gave the model of car and engine type and gearbox
Except if I try and access the same domain using curl, it works fine. For some reason, Docker specifically is what's failing.
Different programs have different defaults.
But in your situation which would be more helpful - prevent this one docker command from using ipv6 (likely more difficult), or preventing all commands from using your broken ipv6 config (likely easier)?
I have no idea about the first. Maybe some people know this detail. But I'm sure that with a distro and version that you're running, there are lots of people who could help with the second. Raspberry Pi 3B+ is the hardware. What software are you using?
Whatever the latest version of Raspbian was a month or two ago when I installed it. uname -a outputs [...]6.12.25+rpt-rpi-v8[...]. /etc/os-release contains "Debian GNU/Linux 12 (bookworm)".
Ok, so it's probably using NetworkManager. I would try disabling it in /etc/NetworkManager/NetworkManager.conf by adding a block like:
[ipv6]
addr-gen-mode=stable-privacy
method=disabled
Then sudo systemctl restart NetworkManager.
Can't say for sure if this will work. I dislike using NetworkManager on my servers so I can't test if this works. But hopefully the before/after of ip addr is different.
Although it looks like your ip addr output posted an hour or so ago doesn't show any ipv6 addressing. Maybe the problem is solved now.
Unfortunately not.
docker run hello-world
Unable to find image 'hello-world:latest' locally
docker: Error response from daemon: Get "https://registry-1.docker.io/v2/library/hello-world/manifests/sha256:ec153840d1e635ac434fab5e377081f17e0e15afab27beb3f726c3265039cfff": dial tcp [2600:1f18:2148:bc00:eff:d3ae:b836:fa07]:443: connect: network is unreachable
Run 'docker run --help' for more information
Well crap. Do you have no ipv6 address now in ip addr?
Guess I gave Docker too much benefit of the doubt and assumed it should failover to v4 once v6 was disabled. Bad assumption on my part.
Could it be a DNS problem? If you dig registry-1.docker.io +short does it return an ipv4 or v6 address?
It looks like there have been sporadic reports of problems from people since last year.
Try adding {"ipv6": false} to your /etc/docker/daemon.json file (create it if it doesnt exist), then restart docker with sudo systemctl restart docker - this forces docker to use IPv4 only.
This flag seems to only disable ipv6 on the default Docker bridge network, not daemon-wide. At least per this discussion.
Do you have no ipv6 address now in ip addr
Just comparing it by eye, there's no change.
zag@raspberrypi:~ $ man dig
No manual entry for dig
zag@raspberrypi:~ $ which dig
zag@raspberrypi:~ $ sudo apt install dig
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
E: Unable to locate package dig
But if I ping it
$ ping registry-1.docker.io
PING registry-1.docker.io (107.20.112.188) 56(84) bytes of data.
It's in the dnsutils package.
52.207.69.161
52.21.128.203
3.224.238.37
44.218.153.24
54.210.249.78
18.232.25.125
52.72.195.97
54.161.169.181
Check your DHCP Server and disable DHCPv6. The raspi shouldn't get a ipv6 then and you won't have this problem in the future with other devices.
edit: and restart the pi after disabling DHCPv6
DHCP server as in on my router?
Still get exactly the same error.
Looks like your Pi thinks it can use IPv6. Find where, and disable it.
Yes, that is what I am here seeking help with.
Absolutely Disable ipv6 - Raspberry Pi Forums https://share.google/ynTkPxr9LkSfz5CZM
Sorry, maybe I should have explained every single step I tried in the body of the original post. I didn't, partly because there were just too many and partly because I forgot which methods I had tried and which pages I had visited. But that page is one I've seen already.
It mentions to "blacklist" the module in a file that doesn't exist. In fact, in a file that's in a directory that doesn't exist, which makes me very sceptical about the later claim that creating that file will fix it. /etc/modules-load.d exists, but not /etc/modules.d. I did already try the final suggestion to add that line to the end of /boot/firmware/cmdline.txt though. No luck.
I think the fact that it's 6 years old is probably a big part of the reason it didn't help. Files and directories have moved around. The suggestions in that post are literally just "do this" without any underlying explanation of what it's doing that could lend to further investigation of the more modern way to do it.
What version of Pi OS are you running?
(we can move to DMs if you wish, it will be less polluting, or stay in the comment thread if you don't feel safe in DMs)
I'm happy to keep it public if only for the off chance that if we find a solution it might some day help someone else with the same issue. The thread'll fall down in the rankings naturally over time anyway so I wouldn't worry about polluting anything for people not actively seeking it out.
I'm not 100% sure how to find the OS version, but uname -a outputs [...]6.12.25+rpt-rpi-v8[...]. /etc/os-release contains "Debian GNU/Linux 12 (bookworm)". It should be whatever was latest as of a month or two ago when I installed the OS fresh.
I have read the rest of the comments to see what you already tried. I was about yo tell you to use sysctl to disable ipv6 but it looks like that is already done.
As a matter of fact, it looks like you have no ipv6 address at all. Which makes me think that your DNS config might be off, as it shouldn't even attempt an ipv6 resolution.
Can you show me the content of /etc/resolv.conf ?
Also install "dig" if you dont have it already and show the output of
dig registry-1.docker.io
$ dig registry-1.docker.io
; <<>> DiG 9.18.33-1~deb12u2-Debian <<>> registry-1.docker.io
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 50801
;; flags: qr rd ra; QUERY: 1, ANSWER: 8, AUTHORITY: 0, ADDITIONAL: 1
;; OPT PSEUDOSECTION:
; EDNS: version: 0, flags:; udp: 1232
;; QUESTION SECTION:
;registry-1.docker.io. IN A
;; ANSWER SECTION:
registry-1.docker.io. 33 IN A 54.210.249.78
registry-1.docker.io. 33 IN A 44.218.153.24
registry-1.docker.io. 33 IN A 107.20.112.188
registry-1.docker.io. 33 IN A 34.234.222.10
registry-1.docker.io. 33 IN A 34.195.83.243
registry-1.docker.io. 33 IN A 52.21.128.203
registry-1.docker.io. 33 IN A 52.0.248.137
registry-1.docker.io. 33 IN A 52.207.69.161
;; Query time: 47 msec
;; SERVER: 192.168.20.1#53(192.168.20.1) (UDP)
;; WHEN: Tue Aug 12 22:27:45 AEST 2025
;; MSG SIZE rcvd: 177
$ cat /etc/resolv.conf
# Generated by NetworkManager
search Home
nameserver 192.168.20.1
edit: oh, and in my router's configuration:
- Primary DNS Server:9.9.9.9
- Secondary DNS Server:1.1.1.1
Well this is getting weird.
Have you tried checking if your os has a resolution cashe active? If so, try to flush it.
I never set one up myself, and this 11 year-old thread implies that Raspbian doesn't have one set up by default. It is a very old thread though, so could be wrong.
edit: this is more recent and also says Raspbian doesn't have DNS caching by default
At this point I am assuming that it is actually a docker issue.
Can you show your docker daemon configuration?
Hard to tell where it is on your machine. Try ~/.docker/daemon.json, or maybe /etc/docker/daemon.json... Else look for it haha
There's a ~/.docker/config.json. In that there're some auths, with keys https://index.docker.io/v1/, https://index.docker.io/v1/access-token, and https://index.docker.io/v1/refresh-token, and then there's "currentContext": "rootless".
There's ~/.docker/contexts/meta/[a long hex string]/meta.json, with {"Name":"rootless","Metadata":{"Description":"Rootless mode"},"Endpoints":{"docker":{"Host":"unix:///run/user/1000/docker.sock","SkipTLSVerify":false}}}.
The only file in /etc/docker is key.json.
It’s transparently pulling from both at the same time and the first one to connect wins. It should just work. Unless you got a typo.
Even if it has an IPv6 address that can't route, it's odd that the OS isn't falling back to IPv4.
But disabling IPv6 in the OS should fix it, as this is not a docker problem.
But disabling IPv6 in the OS should fix it
I've already followed the instructions I could find to do that, but it didn't work. Either disabling IPv6 isn't the solution, or the instructions for how to disable it are incomplete. (You can see comments elsewhere in the thread for more detail.)
Can you give us the full output of the following commands?
ip addr
sysctl net.ipv6.conf.all.disable_ipv6
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
link/ether b8:27:eb:93:9b:06 brd ff:ff:ff:ff:ff:ff
inet 192.168.20.40/24 brd 192.168.20.255 scope global dynamic noprefixroute eth0
valid_lft 81876sec preferred_lft 81876sec
3: wlan0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast state DOWN group default qlen 1000
link/ether b8:27:eb:c6:ce:53 brd ff:ff:ff:ff:ff:ff
4: docker0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue state DOWN group default
link/ether ce:3e:0d:9b:fa:14 brd ff:ff:ff:ff:ff:ff
inet 172.17.0.1/16 brd 172.17.255.255 scope global docker0
valid_lft forever preferred_lft forever
net.ipv6.conf.all.disable_ipv6 = 1