Open Spirit328 opened 5 years ago
Do we have any updates on this? I suffered this issue in CentOS7 too. Could directly run docker commands, while failed in minidcos docker doctor command. Appreciate if any kindly help, thanks.
I could successfully run create action with specifying docker version, command as:
minidcos docker create --docker-version 17.12.1-ce ./dcos_generate_config.sh --agents 0
ref to https://github.com/dcos/dcos-e2e/issues/1252#issuecomment-409343923
although minidcos docker doctor
still failed, the cluster seems works well.
full workround to successfully visit mesosphere web ui page, the command should be run with non-root user:
minidcos docker create --docker-version 17.12.1-ce ./dcos_generate_config.sh --agents 0
minidcos docker wait --cluster-id default
minidcos docker web
Exact same issue with me too.
Host OS: CentOS Linux release 7.6.1810 (Core)
Docker Version: 18.09.1
$ ls -l /var/run/docker.sock
srw-rw---- 1 root docker 0 ජන 12 10:07 /var/run/docker.sock
minidcos --version
minidcos, version 2019.01.10.0
Just like @dennyx said, I also can create and setup the cluster. But still minidcos docker doctor fails.
It would be great if I can get some help with this?
Update: Tried downgrading docker to docker 17.12.1-ce but no joy!
I cannot get miniDC/OS release 2019.05.03.0 to work on CentOS 7 in Docker. However, I can successfully create a cluster with release 2019.05.23.1, CentOS Linux release 7.6.1810, Docker version 18.09.6. However, I do need to specify the Docker version, and minidcos docker doctor
still fails.
To create the cluster:
$ minidcos docker create --variant oss --agents 1 --cluster-id default --docker-version 17.12.1-ce dcos_generate_config.sh
$ minidcos docker wait --cluster-id default
Thanks all for your contributions to this thread. It is interesting and it includes multiple issues:
Traceback (most recent call last):
File "site-packages/tqdm/_tqdm.py", line 931, in __del__
File "site-packages/tqdm/_tqdm.py", line 1133, in close
File "site-packages/tqdm/_tqdm.py", line 496, in _decr_instances
File "site-packages/tqdm/_monitor.py", line 52, in exit
File "threading.py", line 1053, in join
This traceback should no longer be shown, since an update to tqdm
a while back.
2018-12-23 19:19:36 ERROR dcos_e2e._common | docker: Cannot connect to the Docker daemon at unix:///var/run/docker.sock. Is the docker daemon running?.
ERROR:dcos_e2e._common:docker: Cannot connect to the Docker daemon at unix:///var/run/docker.sock. Is the docker daemon running?.
2018-12-23 19:19:36 ERROR dcos_e2e._common | See 'docker run --help'.
ERROR:dcos_e2e._common:See 'docker run --help'.
While folks here have been inspecting their local Docker instances, the error refers to Docker on the minidcos
nodes (Docker in Docker).
--docker-version
In #1574 (released in 2019.06.07.0
), I made the change "Changed the default version of Docker installed on minidcos docker clusters to 18.06.3-ce
.".
This should mean that folks no longer have to specify --docker-version
by default in the newest minidcos
.
What seems clear is that using Docker version 1.13.1 on nodes is problematic on some machines.
The main pain should be taken away with the update to make 18.06.3-ce
the default version.
However, the doctor command will still fail.
Ideally we can narrow down exactly what the problem is, and make that clear in the doctor
error.
An intermediate step might look for the given error and just move on with a warning, potentially one which links to this issue.
After I installed minidcos on my most actual CentOS 7.5
(uname -a: Linux <server> 3.10.0-862.14.4.el7.x86_64 #1 SMP Wed Sep 26 15:12:11 UTC 2018 x86_64 x86_64 x86_64 GNU/Linux)
with the command for the Linux package:sudo curl --fail -L https://github.com/dcos/dcos-e2e/releases/download/2018.12.10.0/minidcos -o /usr/local/bin/minidcos && sudo chmod +x /usr/local/bin/minidcos
I also changed the owner of the directory and the file with:
sudo chown myuser:myuser
... as I have to use sudo on my sytems to install packages. After installation has been finished, I executed the commandminidcos docker doctor -v
All checks went well until 13/13. This gave me the following error:
Actually the docker daemon is up and running:
I tried the same with starting the docker daemon manually:
Then I retried:
on "dockerd" terminal it said:
and the output in the "minidcos" terminal still was:
But a:
seems to proof, that the docker socket is available.
The current user is member of the usergroup "docker" I can execute any docker command without "sudo" upfront. I tried the same commands while being root, but nothing changed.
P.S.: