I recently created a docker-compose file here: https://github.com/ffMathy/Screeps.Server.Docker/blob/master/docker-compose.yml
I'm currently on Windows, but running Linux containers.
When I try to start up my containers using docker-compose up, they appear just fine. However, the screeps container doesn't show anything in the log, and doesn't start (I know this because I can't connect to its port).
Now, if I then click the volume via Kitematic (as shown in the screenshot) and click "Enable volumes" in the prompt that shows, the container restarts, and works - logs are also present in the preview as expected.
I don't understand how this can affect the container itself - that doesn't make much sense to me.
I'm having the same exact issue.
Downgrade Kitematic to 0.17.6 seems to fix the problem.
You can download it from here:
https://github.com/docker/kitematic/releases/tag/v0.17.6
I solved the problem by forcing the volume to mount to a relative path in my docker-compose.yml file.
Related
I have Docker Desktop on Windows 10 and I run Apache NiFi 1.19 within container.
NiFi keeps restarting, by itself, without giving any useful log message, exception message, or whatever I can trace back with.
Any ideas what can be going wrong? I have tried many things.. including this, but nothing helped out.
Ok, finally, so the problem was built-in Hyper-V backend of the Docker (i.e. the default VM environment that comes with Docker Desktop).
I just opted for Use the WSL 2 based engine choice and problem went away.
From the Docker Desktop dashboard, go to settings and in the General tab, select "Use the WSL 2 based enginge":
I saw it but I can't remember where I saw it.
It looked like seeing inside Linux, mounting the docker container on VS Code on Mac OS.
Is it possible to mount the docker container on VS code?
By "mount", I presume you're talking about opening VSCode inside the container instead of Docker volumes and bind mounts. You can install the Remote Containers extension to do this.
Also see the docs mentioned by #AttilaViniczai for how to create development containers/dev contianers for VSCode: https://code.visualstudio.com/docs/remote/create-dev-container
Install the VSCode extension
Select the Remote Explorer tab on the left side of VSCode and make sure you have Containers selected from the drop down menu.
Double click the container you want to work on and/or right click and select Attach to Container. This will attach VSCode to that Docker container and install its required tools inside automatically.
For more information about this, you can see the VSCode documentation here and a tutorial here.
This issue is really strange, but I've no idea how to resolve it. I'm running Win10 and I've got WSL2 on it. When I start Docker with it using the "WSL 2 Based engine", it retrieves some deleted folders out of nowhere. I delete them every single time and after I've restarted my PC and started up Docker again, the folders are back.
There are absolutely no docker containers running ("docker ps" returns nothing) so it couldn't be that some rogue volume definition is being ran along with some container. The folders also only appear once I start up Docker.
The directory inside of which the zombie folders appear is also the source for a Mutagen volume when the containers are running, but as I said - no containers are running.
I ran into the same problem and resolved it by running Troubleshoot from Docker for Windows and purge data on WSL2.
I've not found a solution to selectively clear just that one folder that keeps getting raised from the dead.
I'm currently using vscode's remote containers extension with a .devcontainer.json file that points to my docker-compose.yml file.
Everything works fine and my docker-compose start command gets run (which launches a web server), but I haven't found a way to quickly see the logs from the web server. Has anyone found a way to view the docker log output automatically once vscode connects to the remote container?
I know as an alternative I could remove my container's start command and, after vscode connects, manually open a terminal and start the web server, but I'm hoping there's an easier way.
Thanks in advance!
I'm not using remote containers, just local once, so not sure if this applies but for locally running containers, you can go to the "Docker" tab (you need to install the official Microsoft Docker VS Code Plugin) where you can see your running containers. Just right-click on the container you want to see the logs for and select "View Logs":
You'll see a new "Task" appear in the Terminal pane that will show all your docker logs:
This question is really old and I'm not sure it this option was available at this time but just open the Command Palette (F1) and select/find "Remote-Containers: Show Log".
You see now the log of your container in the terminal.
You can open the command palette and search for: Remote Explorer: Focus on containers view. You should see a sidebar of containers, if you right click your container you can view logs.
I use VS Code's builtin terminal to see the live logs of the docker container that is connected with VS Code.
When VS Code is connected to the docker container, you can open the builtin terminal using the View > Terminal menu option. You should see an existing terminal labeled Dev Containers.
Maybe this is too late? But for others, this is how I do it.
First, instead of logging stuff to the stdout, I redirect all of the outputs into one single file and then using the tail command to steam the output to the terminal instead.
For example, I am going Go here:
logFile, err := os.OpenFile(logFileName, os.O_WRONLY|os.O_CREATE, 0755)
if err != nil {
log.Fatal("Fail to open the log file")
}
logrus.SetOutput(logFile)
Once that's done, I open up my terminal and run my the following command:
$ tail -f {logFileName}
That's one way to do it I guess, but I sure hope VSCode can come up with a better solution.
In the Remote Explorer tab you can see all your docker containers. Under "Dev Containers" is the container for the service specified in devcontainer.json; the rest are in "Other Containers." Simply right click on the container you're interested in and click "Show Container Log." You'll see the full output of the command for that service, just like in an interactive terminal - not a docker build log!
Note I am using a local development container and did not test with remote containers but I'm guessing it's the same.
I have the latest Docker for Mac installed, and I'm running into a problem where it appears that docker-compose up is stuck in a Downloading state for one of the containers:
± |master ✗| → docker-compose up --build
Pulling container (repo.io/company/container:prod)...
prod: Pulling from company/container
somehash: Already exists
somehash: Already exists
somehash: Already exists
somehash: Already exists
somehash: Pulling fs layer
somehash: Already exists
somehash: Already exists
somehash: Downloading [=================================================> ] 234.6 MB/239.3 MB
somehash: Download complete
somehash: Download complete
^^ this is literally what it looks like on my command line. Stopping and starting hasn't helped, it immediately outputs this same output.
I've tried to rm the container but I guess it doesn't yet exist, it returns the output No stopped containers. --force-recreate also gets stuck in the same place. And perhaps I'm not googling for the right terminology but I haven't found anything useful to try - any pointers?
I just needed to restart Docker.
Linux users can use sudo service docker restart.
Docker for Mac has a handy button for this in the Docker widget in the macOS toolbar:
If you happen to be using Docker Toolkit try docker-machine restart.
I faced the same problem! Restarting the service didn't help, downloading again didn't help. It used to get stuck at random instances leaving me with no option but to kill the pull request.
One thing which worked for me was to download 1 file at a time. For Ubuntu users, you can use the following steps:
Stop the service:
sudo service docker stop
Start docker with max concurrent download set as 1:
sudo dockerd --max-concurrent-downloads 1
Download the required image:
sudo docker pull <image_name>
Download images, after that stop the terminal and start the daemon again as it was earlier.
sudo service docker start
I had the similar situation this morning where my network suddenly went down and I was forced to power cycle the modern, while docker-compose was still in the middle of downloading stuff from docker hub.
Yes, bouncing the docker daemon process seems to resolve this.
For Linux users - do sudo service docker restart to fix it.
Go to the Docker Preferences from its menu bar icon. Within there is a "bug" icon. Click on that and then "clean / Purge data"
I'm running OSX and restarting Docker for Mac didn't help. Neither did a full restart or upgrading VirtualBox. What did work was turning my wifi interface on and off every time it got stuck. I had to do this repeatedly, but it eventually downloaded the entire image.
Directly download the necessary images using docker, e.g.
docker pull company/container
and then run
docker-compose up
again. Worked for me on MacOS.
I found a possible workaround.
I have my docker engine installed in a Ubuntu 18.04 Snap Environment.
I discovered searching in some forums that users relate this behaviors to limitation in the download bandwith.
So in the picture below you are going to watch that the components was stucked
Part of the Downloads stucked and finally I cancelled the process CTRL + C
I added two parameters or flags in the configuration file that controls the docker daemon behavior: max-concurrent-downloads 1 and max-concurrent-uploads 1
In my case remember, i am working in a snap environment. This file is located in this directory: /var/lib/docker/current/config/daemon.json
REMEMBER TO STOP ALL DOCKER PROCESS BEFORE THE FILE MODIFICATION, AND CREATE BACKUP OF THE FILE
Add the two lines in the picture. This is going to help you to limit the downloads to only one by one
This is the process that helped me to resolve this problem.
Download Succesfull
I had this issue in my VirtualBox when doing a docker pull on the image but it got stuck at a specific position and never moved from there. So, the issue was due to the network adapters in my VM. I was using NAT by default. When I switched it to "Bridged adapter", the issue went away.
I had a similar problem on docker for windows for a couple of days and when I tried to connect to the virtual machine (via Hyper-V Manager) the downloads started speeding along. I have no idea why but it worked for me...
Completely remove docker
Install docker again
It should work now
I tried to restar docker, update docker, but didnt help