Docker Desktop pull fail with "unexpected eof" - docker

I´m facing the Problem that I can´t pull some Docker images and docker login also works just sometimes.
This occurs only under Windows 10 Pro. I also tested it under Linux and it worked with default configuration.
I´ve already searched for the solution here, but my network connection isn´t the problem, since Linux works on the same network and I also don´t use any Proxy Server which I have to configure.
I´m trying to execute docker pull postgres. The output is:
Error response from daemon: Get https://registry-1.docker.io/v2/library/postgres/manifests/latest: Get https://auth.docker.io/token?account=unkownuser&scope=repository%3Alibrary%2Fpostgres%3Apull&service=registry.docker.io: unexpected EOF
Other images fail also with the same error and sometimes even the docker login fails with the unexpected EOF error.
Everything worked fine since the last update.
I´ve already downgraded my version, but it didn´t help. I also tried to switch from windows to linux containers, no effect.
Does anybody else have this problem or did I do something wrong?

Do you use a FritzBox?
If yes try the following it helped me:
Log into your FritzBox
In the bottom left Corner is a Button called Content (I assume its called that. I got the German version which says "Inhalt")
Click "FRITZ!Box Support"
Disable Hardware acceleration

Related

docker won't start! I'm stuck on "Docker Desktop stopped

I've tried installing Docker and so far I've never been able to use it on my Ubuntu machine:
when I try to start Docker Desktop it only shows Docker Desktop stopped... for an indefinite amount of time. There sometimes shows up randomly an Alert with the messages "Unable to calculate image disk size" for a second or so, but it doesn't really help me at all. I'm also not Using Docker very often so I have only a little experience with it.
At this point Docker is always the first program that shows up automatically when I boot up my computer. and it instantly goes to the message: "Docker Desktop stopped..." when I click on the bug Icon I already tried the oprions "Restart", "Clean/Purge data" and "Reset to factory defaults" but none of those options helped me so far. I also can't access the Settings page it only shows a loading animation when I try to open it.
I have the same problem. The only solution I found is to reinstall docker every time I log in. It seems to work fine until I restart the computer.
It is a very bad solution but it's the best that I've got so far.

Docker Desktop shows error on Windows; It looks like there is an error with Docker Desktop

Windows 10 (19043.1889)
Docker Desktop 4.11.1
I installed docker but it showed "It looks like there is an error with Docker Desktop, restart it to fix it" when opening [Settings] page
I couldn't change any params in [Settings]; I can check some configurations but [Apply & Restart] button didn't enabled.
Several times I tried uninstall & install, but nothing changed.
Docker itself worked, but because of error, I couldn't change resources spec.
I wonder if someone has any idea
Thank you

Docker desktop / daemon won't start on new M1

When I try to startup Docker Desktop on a new M1Max, it gets a fatal error that the "Docker daemon failed to start". The last log item for the daemon is just a debug message "Calling HEAD /_ping".
I'd like to try to start the daemon manually on my Mac to see if there are better error details, but all of the docs for that mention using dockerd which isn't available on the Mac install.
Background: I just switched from an older MBP to a new M1Max. I replaced my Docker Desktop with the new version for Apple Silicon and everything started up fine. However, some of my containers were having issues on the new arch. While debugging those issues, I restarted Docker Desktop and began experiencing this issue. I've tried restarting my mac, but it just gives me the same error again. I've tried downgrading Docker Desktop to the last handful of older versions and still get the same message. The strange thing is that my log file for dockerd hasn't been touched since before I restarted Docker Desktop when I first encountered the error. It's like the daemon isn't even trying to start anymore. Any help would be greatly appreciated.
UPDATE: I just reset my Mac back to my starting point (after using the migration assistant to port my old MBP to the new). When I first install Docker Desktop for Apple Silicon, it starts up normally and everything runs fine. If I check on the dockerd logs, I can see it being written to. However, as soon as I restart Docker Desktop, I get the error mentioned above and dockerd stops writing to logs. Then, every time I try to start Docker Desktop, it errors with the message above.
do the following
clean uninstall docker either using brew/appcleaner
Manually install 4.5.0 M1 - https://docs.docker.com/desktop/mac/release-notes/
that is after manually installing ensure to uncheck automatic software update from setting
!! It worked for me !!

OAuth error when attempting to start-host for chrome-remote-desktop on GCE

I'm running CentOS 7 in a Google Compute Engine VM, and trying to connect to it via Chrome Remote Desktop. When I run the start-host command, it fails with "OAuth error" and no more information.
I accessed the https://remotedesktop.google.com/headless site with Chrome, filled out the choices, and went through the authorization phase. There was not an option for CentOS, so I chose "Debian" and when it gave me the connect command, I replaced the path to start-host with the CentOS one of /usr/lib64/chrome-remote-desktop/start-host, and running the resulting command fails.
I did the likely web searches and found some fairly old and unhelpful information, but nothing useful.
If I can't get Chrome Remote Desktop working, I'd be willing to try another method to run GUI-based applications (in this case, IBM's oneWEX). I can get X11 to sort-of work remotely, but not well enough to run oneWEX.
This is the command I got from the Chrome site, modified to work with CentOS:
DISPLAY= /usr/lib64/chrome-remote-desktop/start-host --code="<code>" --redirect-url="https://remotedesktop.google.com/_/oauthredirect" --name=
The start-host command asked the expected questions of the computer name and PIN, then gave the error:
Couldn't start host: OAuth error.
When you get the code from https://remotedesktop.google.com/headless it is only good for a few minutes, and only one time use.
Just go back to https://remotedesktop.google.com/headless follow the prompts and get a new code for every attempt.
In my case, I got another error when I first ran the command. I tried to resolve it, and reran the same command (with the original code). I kept getting the "Couldn't start host: OAuth error." until I figured out I needed to get a new code from https://remotedesktop.google.com/headless for each attempt.
FYI: The error message is not given right away, but after entering the computer name and double entering the pin.
According to https://bugs.chromium.org/p/chromium/issues/detail?id=1039016, this could apparently also happen if the /usr/lib64/chrome-remote-desktop/start-host binary is not one from the "official" DEBian package built and distributed by Google, but one built from source e.g. by yourself or your Linux distribution (such as the chrome-remote-desktop package built by Fedora; https://bugzilla.redhat.com/show_bug.cgi?id=1788448 aims to clarify how to use that one).
What I can see here is that you previously added some libraries to your CentOS instance, I tried replicating the error myself and inside my CentOS 7 instance (from Compute Engine) I don't seem to have any /usr/lib64/chrome-remote-desktop/ folder.
Or is it from your local machine?
I had the same error and found that refreshing the headless page to get a new code string in the command worked.

Tag latest not found in repository docker.io/library/centos on Redhat Enterprise Linux behind proxy server

I have installed and pulled from docker.io successfully many times on Redhat Enterprise Linux. But now I am having a very difficult time with a slightly different environment behind a proxy server.
I am starting to think there is a modification Redhat has done to the docker packages preventing my proxy setting from working correctly. When I try to do a simple "docker pull centos", I get the frustrating error: "Tag latest not found in repository docker.io/library/centos". I have tried other tags as well with no luck. Before I put the proxy setting in, I received "authorizationrequired" messages. I got past that by adding the proxy info to /etc/sysconfig/docker. Then I received x509 errors. I got past that by adding the appropriate certificate file to docker. Now, with everything seemingly correct, I am unable to do a simple pull. From what I can tell, this is not an issue with stock docker. Is there something else that needs to be done to make the redhat docker package work with a proxy? Any help is greatly appreciated. I tried everything in another stackoverflow question to no avail.
Everything looked like it was working from the cli. However, I ran /var/log/messages | grep docker to look at the docker daemon logs. I was still getting certificate errors. I loaded hub.docker.com into my browser and extracted the certificate chain. I put those in /etc/docker/certs.d/docker.io and it worked part-way.
I needed to add those proxy certs to the global location and now it works all the way.

Resources