I have just installed docker using docker-toolbox 1.8.2 on Windows 10.
Due to due to this issue I had to recreate the docker image using these commands
docker-machine rm default
docker-machine --native-ssh create -d virtualbox default
After that it has been working fine, except for one problem:
When the PC has gone to sleep and then wakes again, the docker commands can no longer connect. Example:
> docker images
An error occurred trying to connect: Get https://192.168.99.100:2376/v1.20/images/json:
dial tcp 192.168.99.100:2376: ConnectEx tcp: A connection attempt failed because the
connected party did not properly respond after a period of time, or established connection
failed because connected host has failed to respond.
However the docker-machine lists the machine as running:
> docker-machine ls
NAME ACTIVE DRIVER STATE URL SWARM
default * virtualbox Running tcp://192.168.99.100:2376
I can also confirm in VirtualBox that the VM screen seems to be active.
I have tried starting and stopping the machine, but that does not help
C:\x> docker-machine stop default
C:\x> docker-machine start default
Starting VM...
Started machines may have new IP addresses. You may need to re-run the `docker-machine env` command.
C:\x> docker-machine env default --shell=powershell
Ironically, the last command hangs, so I never get any environment settings.
The only thing that helps is to restart the whole PC. But that should be unnecessary?
I have also posted this as an issue on the docker github repository,but that was closed. A related issue seems to be this one, but no workaround or solution has been posted for Windows.
After hous of fighting with VirtualBox + Docker Toolbox, I finally found the way, how to make Docker working again (even without restarting all the containers):
Wake up PC from sleep
Try docker images (won`t work)
Open VirtualBox -> Close VM with saving state (CTRL+V)
Run your VM again
Try docker images again (now should work)
Please note: All steps are in VirtualBox only! Running docker-machine restart default will create another host-only adapter, which is something you do not want. If you did it anyway, delete all additionally created adapters (File->Preferences->Network on VirtualBox), then follow steps 1-5.
I have experienced the exact same symptoms on Windows 8.1... The thing is that it's not really a docker-specific issue, but more how Windows manages the VirtualBox network adapters after sleep (I think...). The culprit in my case is that the network adapter's addresses were becoming private after sleep (they became 169.* addresses).
Credits to this guy who gave me the idea: http://lyngtinh.blogspot.ca/2011/12/how-to-disable-autoconfiguration-ipv4.html
Fix:
Start a command prompt as Administrator
Find out the "useful" network adapters: ipconfig /all. The useful ones in my case were the ones labeled "VirtualBox Host-Only Ethernet Adapter" that didn't have private ips (not starting with 169.*).
Run this command and note the "Idx" of the useful VirtualBox network adapters: netsh interface ipv4 show inter.
Run this command to disable the IP auto configuration: netsh interface ipv4 set interface <idx> dadtransmits=0 store=persistent. Replace <idx> with each index found in the previous step.
Restart Windows
Afterwards, I was able to docker-machine start default, then docker-machine env default --shell cmd, put the PC to sleep, wake up and run docker-machine env default --shell cmd again.
I found that removing 'host only adapter' (File->Preferences->Network on VirtualBox), and restart the docker-machine helps.
Not a real solution. But probably better over restart the computer.
Having tried all the other answers here, and having varying but not consistent success, the following seems to reliably bring it back for me after this problem occurs.
Open a powershell/command window (I have most success if I run all docker-machine commands in a powershell window opened as administrator, I don't know if that's important or not) then run (where "dev" is the name of your docker machine instance):
docker-machine ssh dev
Then on the terminal that is opened, run:
sudo shutdown -r now
When the machine restarts, it seems to refresh the network and work correctly. Note, however, that simply running docker-machine restart dev did not have the same effect for me.
Your machine needs to be running before you can do the ssh, so if it's not running, execute docker-machine start dev before trying to SSH.
Had the same problem on Windows 8.1 and docker toolbox 1.12.0
None of the above solutions worked for me, too.
[edited]
Found another way to make docker work after system wake up:
In the docker Quickstart Terminal window, stop docker process Ctrl-C (if it is still running)
Run command docker-compose down
Shut down docker with docker-machine stop default
Exit terminal window Ctrl-D
Run Quickstart Terminal again and do all subsequent steps you need.
This worked for me, on Windows host machine.
Configure your network adapter to
1) Allow the network adapter to wake the computer,
2) Allow a magic packet to wake the computer,
3) Allow IPV6
http://www.worldstart.com/dropped-internet-connection-in-sleep-mode/
Also, on virtual box network settings, go to advanced, and allow promiscuous mode to VM machines, or allow all
Related
Original Post
I have a Windows workstation with WSL2 and Docker installed that I am able to use for container based development in VS Code. I would like to be able to develop inside the containers on this system remotely. I am able to SSH directly into the WSL2 environment on the workstation and am able to start the docker daemon without logging directly into Windows by creating a Task to start the daemon automatically as described here: https://stackoverflow.com/a/59467740/10692741
However when I try to access Docker on the remote machine by following this guide: https://code.visualstudio.com/docs/remote/containers-advanced#_developing-inside-a-container-on-a-remote-docker-host, I get the following error:
error during connect: Get http://docker/v1.24/version: net/http: HTTP/1.x transport connection broken: malformed HTTP status code "\x00c\x00o\x00m\x00m\x00a\x00n\x00d\x00"
I have also tried connecting via a SSH tunnel as outlined here: https://code.visualstudio.com/docs/remote/troubleshooting#_using-an-ssh-tunnel-to-connect-to-a-remote-docker-host and am unable to connect to Docker as well.
Has anyone had success with a setup like this? Or is this not supported due to limitations with Docker on Windows, WSL2, and/or Windows OpenSSH implementation?
Update: 2021-01-21
When I SSH into the Windows machine remotely, I am able to see the docker containers in the VS Code extension. I am able to start them, stop them, and enter into them with the shell. However, when I try to attach VS Code I get same error shown above.
Things that may have possibly affected this over the past couple days:
Adding SSH keys on my local machine to the ssh-agent via ssh-add /my/key
Exposing Docker daemon on tcp://localhost:2375 without TLS on the remote Windows machine
Also I want to note that the I've tried using Windows, Mac, and Linux as the local machine. With Mac and Linux I am able to open a remote session into the Windows machine, but from the Windows local machine I am able to SSH into the remote Windows machine but cannot open a remote connection in VS Code for some reason.
Ok, I was able to get this working using the port/socket forwarding technique. For sake of clarity, I'll use:
local development workstation, local workstation, or just workstation to indicate the computer from which we wish to use VSCode to access Docker containers on ...
the remote Docker host, remote, or just Docker host
Sanity check -- Do you have Docker Desktop installed on both systems? On the local development workstation, you can skip the WSL2 integration, but you'll at least need the client tools, since the VSCode extension uses them.
Steps I took:
I already had Docker with WSL2 integration set up on my main system (which for the purposes of this exercise, became my remote Docker host), along with VSCode, so I knew everything was working there. It sounds like that was your starting point as well.
On another system on the same network (accessed with RDP to make it simple), I already had VSCode installed as well, with the Remote Development Extension Pack. I also have WSL on that system, but only a v1 instance there. Not that WSL on the workstation should be a factor at all for the purposes of this exercise.
I installed Docker Desktop for Windows on that local development workstation.
I also installed the Docker extension for VSCode, since I didn't yet have it on the local development workstation.
On the workstation, I was not yet set up to SSH from PowerShell into my WSL Ubuntu distro on the remote. From PowerShell on the workstation, I generated an ECDSA key (per this and other documents) and added the public key to my authorized_keys on the the remote.
On the workstation, I started the OpenSSH Authentication Service and added the newly created key to the agent (in PowerShell) with ssh-agent add ~\.ssh\id_ecdsa.
I logged out of the workstation and back in so that the path changes were picked up for the Docker desktop install.
I was then able to ssh from Powershell on the local to Ubuntu/WSL on the remote with the port forwarding. Since I'm using the Windows 10 OpenSSH server as a jumphost to my WSL SSH servers, my command looked slightly different (with a -o "ProxyCommand ... mainly), but overall the structure is the same as the one listed in the "SSH Tunnel" doc you linked in your question.
On the remote (manually, not through any integration from the local), I did a basic docker run -it --rm Ubuntu and left it open.
On the local, from PowerShell, I set the DOCKER_HOST environment variable via [System.Environment]::SetEnvironmentVariable("DOCKER_HOST","tcp://localhost:23750").
I was then able to see the remote container using docker ps on the local. I could also docker exec -it containername bash into it remotely.
Of course, the above two steps aren't needed in the long term for VSCode, they were just part of my process to make sure everything was up and running (since, as you might expect, I did have several points at which I failed during this process).
So with that working, it was a simple matter in VSCode to change the Docker extension's DOCKER_HOST setting to tcp://localhost:23750. And voila, I could see all images on the remote as well as attach to them from VSCode.
Other thing(s) to check
I'll add to this list if we find additional reasons why it might not be working, but for now:
You mention that you are starting the Docker Desktop daemon automatically at startup via Task Manager, but you don't mention anything about the WSL2 instance. However, since you are able to ssh into it, I assume you have a way to bring it up as well? My experience has been that, unless the owning user is logged in, WSL terminates any instances after a few seconds, even if a service is running. There's a workaround, I believe, that I can dust off if this is a problem.
I always SSH into my virtual machine and don't use VMWare's virtual machine window/terminal after I am done setting up the VM. I find it annoying to always have the VMWare Fusion application open as well as the VM window, thus, I want to be able to silently start the VM from my terminal and do all other actions only through terminal.
I saw this page on the VMWare website that talks of starting a VM with the command:
/Applications/VMware\ Fusion.app/Contents/Library/vmrun start full/path/of/your/virtual/machine/bundle
That is a step in the right direction as I don't need to click anything to start my VM, however, it still opens the application and VM in windows.
When I used Virtual Box, I remember being able to do this with a command like:
VBoxManage startvm "VM name" --type headless
Is there a similar command that for VMWare that would allow me to run my VM headless?
You can try :
vmrun start full/path/of/your/virtual/machine/bundle nogui
I'm using a docker with Windows and Hyper-v to create containers. I've added a docker machine vmachine to my docker configuration. First time the machine is created, it gets an IP (although I cannot manage nginx to access it - ERR_CONNECTION_REFUSED) and finishes the bootup.
When I turn off the machine and then try to boot it, i get stuck in this message
Waiting for SSH to be available...
And it doesn't evolve from there. The machine is booted, however, I get an IPv6 when I input the command docker-machine ip vmachine like - fe80::215:5dff:fe21:10b insted of a IPv4
What am I doing wrong?
Problem here is by default docker uses DockerNAT network switch. You should create a new external network switch instead. This issue is covered here and here. You can create an external network switch using the below command
docker-machine create -d hyperv --hyperv-virtual-switch external-switch tempbox1
or you can create one through the UI
Be sure to reboot the device after creating the external switch.
I had a similar issue and non of the solutions worked. Turns out that according to this answer, docker launches SSH with Unix specific elements. This is said to have been fixed in the releases that followed, but I still encountered the 'Waiting for SSH' issue. I resolved this by simply using GIT bash to run all docker related SSH commands.
Use the switch --native-ssh
for example docker-machine --native-ssh .... Get more details from here
docker-machine.exe -debug create --driver hyperv --hyperv-virtual-switch "External Virtual Switch" --hyperv-cpu-count "1" --hyperv-memory "1024" --hyperv-disk-size "20000" mydockervm
make sure to have additional VirtualSwitch configure , with external network driver selected , Uninstall virtualbox
Use the debug switch to see the exact error , for me it was not able to allocate memory.
Here's what's solved it for me.
Turns out Windows 10 starting version 1709 has a built in SSH client at C:\Windows\System32\OpenSSH. Here's an article discussing it.
Looks like docker is using that SSH implementation and it's not compatible. I didn't look for a proper way to remove the built-in SSH implementatino in Windows 10, and simply renamed the folder. That was enough to fix it for me.
After doing what is mentioned in the above suggestions and if you are running docker on a windows machine try to login using cli. This has worked for me.
If you are using Command Promt Docker will stuck at Waiting for SSH to be available..., So change to use GIT BASH as #Dave Howson said it will work.
If you're using oracle VM you must ensure first that your new cloud vm is running.
Before:
After:
I am running the latest Docker Toolbox, using latest Oracle VirtualBox, with Windows 7 as a host OS.
I am trying to enable non-TLS access to Docker remote API, so I could use Postman REST client running on Windows and hit docker API running on docker-machine in the VirtualBox. I found that if Docker configuration included -H tcp://0.0.0.0:2375, that would do the trick exposing the API on port 2375 of the docker machine, but for the life of me I can't find where this configuration is stored and can be changed.
I did docker-machine ssh from the Toolbox CLI, and then went and pocked around the /etc/init.d/docker file, but no changes to the file survive docker-machine restart.
I was able to find answer to this question for Ubuntu and OSX, but not for Windows.
#CarlosRafaelRamirez mentioned the right place, but I will add a few details and provide more detailed, step-by-step instructions, because Windows devs are often not fluent in Linux ecosystem.
Disclaimer: following steps make it possible to hit Docker Remote API from Windows host, but please keep in mind two things:
This should not be done in production as it makes Docker machine very not secure.
Current solution disables most of the docker-machine and all docker CLI functionality. docker-machine ssh remains operational, forcing one to SSH into docker machine to access docker commands.
Solution
Now, here are the steps necessary to switch Docker API to non-TLS port. (Docker machine name is assumed to be "default". If your machine name has a different name, you will need to specify it in the commands below.)
Start "Docker Quickstart Terminal". It starts Bash shell and is the place where all following commands will be run. Run docker-machine ip command and note the IP address of the docker host machine. Then do
docker-machine ssh
cd /var/lib/boot2docker
sudo vi profile This starts "vi" editor in elevated privileges mode required for editing "profile" file, where Docker host settings are. (If as a Windows user you are not familiar with vi, here's is super-basic crash course on it. When file is open in the vi, vi is not in editing mode. Press "i" to start edit mode. Now you can make changes. After you made all the changes, hit Esc and then ZZ to save changes and exit vi. If you need to exit vi without saving changes, after Esc please type :q! and hit Enter. ":" turns on vi's command mode, and "q!" command means exit without saving. Detailed vi command info is here.)
Using vi, change DOCKER_HOST to be DOCKER_HOST='-H tcp://0.0.0.0:2375', and set DOCKER_TLS=no. Save changes as described above.
exit to leave SSH session.
docker-machine restart
After doocker machine has restarted, your sould be able to hit docker API URL, like http://dokerMachineIp:2375/containers/json?all=1, and get valid JSON back.
This is the end of steps required to achieve the main goal.
However, if at this point you try to run docker-machine config or docker images, you will see an error message indicating that docker CLI client is trying to get to the Docker through the old port/TLS settings, which is understandable. What was not expected to me though, is that even after I followed all the Getting Started directions, and ran export DOCKER_HOST=tcp://192.168.99.101:2375 and export DOCKER_TLS_VERIFY=0, resulting in
$ env | grep DOCKER
DOCKER_HOST=tcp://192.168.99.101:2375
DOCKER_MACHINE_NAME=default
DOCKER_TLS_VERIFY=0
DOCKER_TOOLBOX_INSTALL_PATH=C:\Program Files\Docker Toolbox
DOCKER_CERT_PATH=C:\Users\USERNAME\.docker\machine\machines\default
the result was the same:
$ docker-machine env
Error checking TLS connection: Error checking and/or regenerating the certs: There was an error validating certificates for host
"192.168.99.101:2376"
If you see a problem with how I changed environment variables to point Docker CLI to the new Docker host address, please comment.
To work around this problem, use docker-machine ssh command and run your docker commands after that.
I encountered the same problem and thanks to #VladH made it working not changing any internal Docker profile properties. All you have to do is correctly define Windows local env variables (or configure maven plugin properties, if you use io.fabric8 docker-maven-plugin).
Note that 2375 port is used for non-TLS connections, and 2376 only for TLS connections.
DOCKER_HOST=tcp://192.168.99.100:2376
DOCKER_TLS_VERIFY=0
DOCKER_TOOLBOX_INSTALL_PATH=C:\Program Files\Docker Toolbox
DOCKER_CERT_PATH=C:\Users\USERNAME\.docker\machine\machines\default
I am new to docker and nowhere near a networking expert, but I am seeing some strangeness when trying to run a docker container instance (right word?). I am running docker on OSX and set it up using the documentation found here: http://viget.com/extend/how-to-use-docker-on-os-x-the-missing-guide
Everything seems to have gone along fine, then I set up the port forwarding rules via these lines:
for i in {49000..49900};
do VBoxManage modifyvm "boot2docker-vm" --natpf1 "tcp-port$i,tcp,,$i,,$i";
done
I can confirm the boot2docker VM instance by checking the configuration within the Oracle VM VirtualBox Manager->Network->Adapter 1->Port Forwarding with the OSX.
I then run this command to get the container.
docker run -d -P dockerhub.emory.edu/ecoi_trunk:2
I do a "docker ps" and get this info.
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
f20bfefa2e97 dockerhub.emory.edu/ecoi_trunk:2 "/usr/sbin/apachectl 18 seconds ago Up 15 seconds 0.0.0.0:49153->443/tcp, 0.0.0.0:49154->80/tcp cranky_einstein
However, when I run an "lsof -i :49153" I see nothing is listening. I also can't reach the container via the "l****host:49153" in my browser. It just hangs.
What's strange is if I explicitly set the port (rather than allowing docker to assign one) via the following command:
docker run -d -p 49000:80 dockerhub.emory.edu/ecoi_trunk:2
It seems to work (lsof -i:49000 displays a TCP LISTEN), and I can confirm it's listening and the container is reachable via "l****host:49000". However, it's extremely slow. I'm not sure if either are related, but would welcome any tips or thoughts.