JupyterHub and BinderHub Team Meeting - October#
Date: Thursday, October 17, 7AM UTC
name | institution | GitHub handle
Erik Sundell | Sandvik | @consideRatio
Tim | Binder | @betatim
Sarah | Turing | @sgibson91
Simon | Open Microscopy Environment | @manics
Kenan | GESIS | @bitnik
Arnim | GESIS | @arnim
Min | Simula | @minrk
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=Erik Sundell] About z2jh progress.
[name=Simon] Thanks for inviting me to join!
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.
[name=Erik] Chia-liang Kao (@clkao) and Simon Li (@manics) have done a lot of work I greatly appreciate regarding Local development and CI/CD in z2jh! :heart:
[name=Sarah] I’ve begun deploying the Turing federation cluster!
[name=myname] My report!
This is what we’ll cover in the meeting (we have about 60m in total). We’ll copy the proposed agenda topics above here just before the meeting.
Add agenda item here [name, estimated time for conversation]
Jupyter Governance changes [name=Tim] (X minutes)
details needed from Tim (check forum thread for details)
several threads including lots of new ones this morning
if you are interested in governance get yourself involved!
mybinder.org cloud credits contingency [name=Tim]
what should we do if we run out of GKE credits?
jupyterhub/team-compass#214 related discussion issue
top repos by launches (look at timestamp column)
launches per repo in the 120 days leading up to 2019-10-01
give less resources to repositories that are used mostly to “just look”
will only work if we are limited by memory not number of pods per node
need to stick to highmem-8 nodes on GKE
TODO: What node types are used across the federation (pods per node limit, memory and CPU)?
this will help determine resource limits.
Turing: Standard_D4s_v3 nodes, 4 vCPU, 16 GB
TODO: implement per repo resource limits
TODO: try out what happens during a version bump when capacity will reduce briefly
Investigate mechanisms to make the “don’t send traffic” more granular for version mismatches
TODO: what are the special services we run on GKE that can’t easily move?
MS Azure Portal walkthrough [name=Sarah] (10 mins)
Tim asked me if I could do a short introduction to the Azure Portal for managing an Azure-hosted BinderHub
Would anyone want a longer meeting for more questions?
Docker Bot for Azure: HelmUpgradeBot/DockerCleanUp
can we setup a shared docker registry that is shared amongst all clusters in the federation?
add option to configure federation members to only opt-in to “high traffic” repositories
Software Sustainability Insitute 2020 Fellowship application run-through [name=Sarah] (10 mins)
I’ve decided to apply to do something Binder related and the application involves a 6 minute screen-cast presentation.
Looks like we may not have time for this so here’s the link to the PR with the script for my presentation, a link to the slides is at the top of the markdown file: sgibson91/ssi-fellowship#1
Releasing JupyterHub 1.1 [name=min]