Ask quick dev questions about mybinder.org, BinderHub, and repo2docker. Use discourse.jupyter.org for user questions, support, and discussion.
Should be fine with flake-compat, I'll probably get round to testing it at some point
Just experimenting with this now in my lunch break: repo2docker is building... 🤞
nix flake update
flake.nix and flake.lock are present, but until then using a shim default.nix like the one in that repo will work fine
binder/ directory, however I don't find any mention of this in https://mybinder.readthedocs.io/. I'm trying to point my students to some proper documentation, am I looking at the right place?
1 reply
1 reply
well, it turns out that it was on the first paragraph of the "what is binder" explanation 🙈 https://mybinder.readthedocs.io/en/latest/introduction.html#what-is-a-binder
Configuration files may be placed in the root of your repository or in a binder/ folder in the repository’s root (i.e. myproject/binder/).
Hello! We are trying to upgrade our binderhub kubernetes helm chart to the latest version 0.2.0-n636.h8e3b5ac from 0.2.0-n611.he777436 . When we try and do this, we get a helm error;
Error: UPGRADE FAILED: values don't meet the specifications of the schema(s) in the following chart(s):
binderhub:
- (root): Additional property etcJupyter is not allowedBeginning at the first line of our config.yaml, it looks like this;
etcJupyter:
jupyter_notebook_config.json:
NotebookApp:
allow_origin: '*'
# shutdown the server after no activity
shutdown_no_activity_timeout: 3600
MappingKernelManager:
# shutdown kernels after no activity
cull_idle_timeout: 3600
# check for idle kernels this often
cull_interval: 60
# a kernel with open connections but no activity still counts as idle
# this is what allows us to shutdown servers
# when people leave a notebook open and wander off
cull_connected: trueI've noticed that mybinder.org also used this same section, and is on the latest helm chart. This latest version added a new sort of scheme, so we suspect this is causing issues for us, but not for mybinder.org somehow. Our chart work perfectly find on version 0.2.0-n611.he777436. Does anyone have an idea about what could be going on?
If you do want something like etcJupyter, which is a way to inject a jupyter_notebook_config, you can use hub.extraFiles part of the jupyterhub Helm chart instead.
See the configuration reference at z2jh.jupyter.org for that.
Alright, we were able to upgrade our helm chart, but the issue we sought to fix is still occurring.
The health endpoint for our binderhub (https://binder.libretexts.org/health) oscilates between {"service": "Docker registry", "ok": true} and {"service": "Docker registry", "ok": false}. at around avg 30 minute intervals (we are using dockerhub). We were hoping the new helm chart would fix this due to a url change, but it did not. This started happening on July 19th.
Is there anyway that we can figure out the exact URL that health.py is using to check the docker registry? That way we can try and solve this.
Hi,
I am writing a jupyter-book with Binder integration. However, since my book is aimed at scientists the data to handle can become very large. The 2GB RAM limit is far too low for my needs. I would be happy to supply a custom runner for my binder instances. Is this possible in any way?
Not sure if this is the right place to ask, but does anyone know if it's possible to make a mybinder link such that the initial thing that's displayed to the user when they click on the link and load the binder is jupyterhub with a terminal open?
I can see how to make a url that opens a file by using something like urlpath=lab/tree/ipynb/SquareSum15.jpg, as in https://mybinder.org/v2/gh/norvig/pytudes/main?urlpath=lab/tree/ipynb/SquareSum15.jpg, but I'm not sure if it's possible to have the urlpath specify that a terminal should be open
1 reply
