Hello,
I am attempting to run a local container of the lastest renkulab-py docker image. I initiated jupyterhub and I can access to the login page without issues. However, when I try to introduce the default password “jupyter” and the user “joyvan” it does not work. Has been the default password modified? If so, what is it the new password?
In other docker images such as renulab-R I have no problem as “rstudio” user can be used with “rstudio” as pasword.
Thanks in advance,
Carlos
Hi @caviri, thanks for reaching out on the forum!
We don’t set a password for the notebook server; instead access works via a token. E.g. I can do something like this
$ docker run --rm -ti -p 8888:8888 renku/renkulab-py:3.9-0.12.0 jupyter lab
...
[I 2022-09-22 13:05:50.940 ServerApp] Jupyter Server 1.15.6 is running at:
[I 2022-09-22 13:05:50.941 ServerApp] http://651f26261152:8888/lab?token=9ea0fbefc1a288fe54d0a3f21162b45a3c02de69a1cd7378
[I 2022-09-22 13:05:50.941 ServerApp] or http://127.0.0.1:8888/lab?token=9ea0fbefc1a288fe54d0a3f21162b45a3c02de69a1cd7378
[I 2022-09-22 13:05:50.941 ServerApp] Use Control-C to stop this server and shut down all kernels (twice to skip confirmation).
[C 2022-09-22 13:05:50.957 ServerApp]
To access the server, open this file in a browser:
file:///home/jovyan/.local/share/jupyter/runtime/jpserver-21-open.html
Or copy and paste one of these URLs:
http://651f26261152:8888/lab?token=9ea0fbefc1a288fe54d0a3f21162b45a3c02de69a1cd7378
or http://127.0.0.1:8888/lab?token=9ea0fbefc1a288fe54d0a3f21162b45a3c02de69a1cd7378
If I copy that last line and put it into a browser, it connects automatically.
If you are doing this inside a Renku project and you have a recent version of the Renku CLI installed, you can also do
$ renku session start
The container image 'test-wf:b3bde6e' does not exists. Would you like to build it? [y/N]: y
Image test-wf:b3bde6e built successfully.
Session The session for 'test-wf:b3bde6e' has been successfully started. It is available at:
http://0.0.0.0:58302/?token=eb5d035660b847208af79692779709e8 successfully started
The session for 'test-wf:b3bde6e' has been successfully started. It is available at:
http://0.0.0.0:58302/?token=eb5d035660b847208af79692779709e8
You can copy-paste the link at the bottom as before, or use
$ renku session ls
renku session ls
ID STATUS URL
---------- -------- ------------------------------------------------------------
108a50ec3e running http://0.0.0.0:58302/?token=eb5d035660b847208af79692779709e8
$ renku session open 108a50ec3e
Hope that helps!
Thanks for the fast response @rrrrrok ! That works!
In my case I wanted to have 2 or more users connected to the hub and I just discover a way to do it. I’ll let it here in case it is useful for somebody. I just added --user root when running docker as this:
docker run -p 8500:8000 --user root --gpus=all --name renkulabTF renku/renkulab-cuda-tf:11.2-tf-2.9-c26fe94 jupyterhub
Then in the CLI you will be logged as root and be able to use adduser XXXXX to add the users you want to have access to the Hub.
Thanks again!
I see now that I probably misunderstood your question - you are running JupyterHub inside the container, not a single session? Are you also using an on-line instance of Renku, e.g. at renkulab.io?
Yes, we are developing a training course based in notebooks at DNF (UNIL) and so far renkulab.io is more than enough for many of those sessions. However I am trying to develop a way to access to GPUs by multiple users at once. For this I am trying to run a JupyterHub (from a modified renku docker image) with a whitelist of users.