JupyterHub and BinderHub Team Meeting - April¶
Date: 11 April 2019 at 8am Zurich time (11pm PST on Wednesday)
Link to Video Recording: click here
The call will start with a quick round of updates from everyone present (you can pass if you want to), followed by bigger agenda items. There will be no feedback/questions on the 60s updates. Questions/discussion on the larger agenda items will be moderated by a speaker list (see end of this hackmd).
Welcome to the Team Meeting¶
If you are joining the team video meeting, sign in below so we know who was here. Roll call:
name / institution / GitHub handle
Zach / Jupyter Cal Poly / @Zsailer
Tim / Binder / @betatim
Min / Simula / @minrk
Kirstie / Turing Institute / KirstieJane
CL / InfuseAI / @clkao
Sarah Gibson / Turing Institute / @sgibson91
Joe Hamman / NCAR / @jhamman
60 second updates on things you have been up to, questions you have, or developments you think people should know about. Please add yourself, and if you do not have an update to share, you can pass.
name: update or pass
Min: JupyterHub 1.0
Tim: Welcome Sarah to the mybinder.org deployer team!
CL: Kind (kubernetes in docker) as faster way to get a kubernetes cluster in CI/laptop
Joe: DevOps all day, every day at the moment (six hubs on multiple cloud vendors!!)
Sarah: Learning how to make mybinder.org deployments and investigating autoscaling on Azure
Reports and celebrations¶
This is a place to make announcements (without a need for discussion). This is also a great place to give shout-outs to contributors! We’ll read through these at the beginning of the meeting.
Let’s collect all potential agenda items here before the start of the meeting (closing time 24h before). We will then attempt to create a coherent agenda that fits in the 60m meeting slot. If there are similar items try and group them already.
Add agenda item here [name, estimated time for conversation]
Configuring the repo2docker base image [Joe Hamman, 5-10 minutes]
hopefully showing off a prototype…
Some good reading on Cloud Native Buildpacks: https://discourse.jupyter.org/t/creating-a-specification-for-reproducible-repositories/569/10?u=betatim
Cloud native buildpacks seem to be the enterprise version of repo2docker
we should understand how the cloud native buildpacks work and see what can be borrowed from their design.
repo2docker’s build pack structure was inspired by heroku buildpacks. We liked their first idea, so maybe we should keep copying them :)
motivation: large images that take a long time to build and then distribute on nodes
motivation: want to share environments between BinderHubs and JupyterHubs
Favoured solution right now based on BuildPacks specifyign their own base image and repo2docker will look for entrypoints so that you can add BuildPacks from other Python packages
can you use a straight up repo2docker image with JupyterHub or not? Should you?
Yes you can
procedure around assigning sub-domains of mybinder.org to other BinderHubs.
decide to accept or reject a procedure if there is a consensus either way
if there is no consensus decide what is missing for people to be able to make a decision (either way)
feedback wanted: PrimeHub - a meta chart for z2jh and keycloak [CL, 5-10min]
motivation: install z2jh and keycloak and allow admin to config auth through keycloak
A meta chart that gives you a working JupyterHub with KeyCloak right away
Asking for feedback:
is this actually useful for anyone?
what do you think about the licensing of the project?
mentioning the various projects it depends on in the README (and LICENSE?) seems like it ticks all the boxes
Adding discourse as a part of this chart is probably out of scope for the current PrimeHub
What timeslot works for the Asia-Europe version of the meeting?
maybe move the meeting back by one hour to make it easier for Europe
clear that if you are based in the Americas “this meeting is too late for you to attend”
create good notes so that the other timezones can catch up
stay with one meetinng per month that alternates instead of creating two meetings per month