Closed CommonClimate closed 2 years ago
@CommonClimate and @khider, the hub is ready for your testing!
It lives in the usual place: https://paleohack2021.hackathon.2i2c.cloud/
@damianavila thanks for this. We were able to use a link to the hub that already has the latest materials pulled from this repo: https://github.com/LinkedEarth/paleoHackathon. The only sticking point we foresee is that the Python environment needs to be updated. We have an environment.yml file to provide but we don't know how to do this on our own. Would it be OK to contact you on Monday when we have our new Pyleoclim version released so you can spruce up the environment for us?
Thanks in advance.
@CommonClimate want to try following the directions in this guide for managing your own Docker image? https://docs.2i2c.org/en/latest/admin/howto/environment.html#customize-your-hub-environment
@CommonClimate could it be the case that you updated/deleted a previous comment? IIRC, the question was about running multiple notebooks... And the answer is yes, you can run multiple notebooks. You also mentioned JupyterLab. If you are interested in that experience, you could configure JupyterLab as the default user interface using the configurator as it is described here: https://pilot.2i2c.org/en/latest/admin/howto/configurator.html#default-interface
Regarding the environment update, as @choldgraf mentioned above, we are trying to follow the linked process now.
But, it is also worth mentioning that your repo already has support to automatically build the image thanks to some preliminary work done by @yuvipanda: https://github.com/LinkedEarth/paleoHackathon/pull/2 So, if you update the environment file in that repository, it should eventually build the new image and push it to quay. Then, it is just a matter to reference the image with the Configurator.
So, in summary, you can self-serve the environment update! And we would recommend that you attempt to do it yourself. But if you need some additional help, we would be happy to assist you.
OK, we'll try to self-serve. Yes, I did delete a comment ; Deborah knew the answer already.
I think I got most of the way there. However, I made a mistake in crating the image name and I'm getting an invalid image error. But can't stop my server (API error).
@CommonClimate I also changed us to JupyterLab
Yep, you had a typo in the image name.
Working
quay.io/2i2c/paleohack-2021:d6234e037f4f
Not working
quay.io/212c/paleohack-2021:d6234e037f4f
Notice the 212c
instead of 2i2c
.
Btw, I have corrected the typo with the configurator. So you should be able to start your notebook server successfully now.
@damianavila Thanks! Yes I made two typos. As I was trying to fix the first one (I left a space between '/' and ' paleo', I couldn't shut down my own server to even try to get it restarted. Any magic trick on that?
And I'm also getting this error a lot:
I can usually manually save after it fails.
Also @damianavila We are going to need 50 nodes for the hackathon.
@khider, since the hub is live and functional we are going to close this issue which was intended for the hub creation. Any further request/question should be tunneled to our support system: https://docs.2i2c.org/en/latest/support.html.
But, let me reply to some of your previous questions before closing this issue 😉 .
Any magic trick on that?
Usually, after 5 min, if the server can not start successfully, it will shut down and you should be able to start it again.
And I'm also getting this error a lot:
Can you please send an email to support about this issue? Thanks!
Also @damianavila We are going to need 50 nodes for the hackathon.
With 50 nodes, you mean you will need to serve 50 people, right? FYI, each person will be served by a pod and those pods are allocated on compute nodes that automatically autoscale when it is actually needed. So, as people start using the hub, new nodes will spin up to serve the upcoming users.
Closing this one now but do not hesitate to contact us through the support channel if you need to!
@khider, FYI, we have pre-warmed the cluster with 5 nodes so your attendees do not have to wait a lot of time for their notebook servers to start up.
@damianavila Thank you But getting this:
2021-10-28T17:28:56Z [Normal] pod didn't trigger scale-up: 1 max node group size reached, 3 node(s) didn't match node selector
Hub Description
Second installment of our Paleoclimate community hackathon, centered around the Python scientific ecosystem for the analysis of paleoclimate observations. The event will bring together 50 participants from around the world for a 2-day virtual event, when they will be working in teams completing exercises coded in Jupyter notebooks utilizing primarily our Pyleoclim package.
Community Representative(s)
Important dates
Target start date
2021-10-21
Preferred Cloud Provider
No preference (default)
Do you have your own billing account?
Hub Authentication Type
GitHub Authentication (e.g., @mygithubhandle)
Hub logo
https://github.com/LinkedEarth/Logos/blob/master/pyleoclim_insigna_light.png
Hub logo URL
No response
Hub image service
No response
Hub image
No response
Extra features you'd like to enable
Other relevant information
There is a precedent as we ran the first version of this event in Feb 18-19 2021 with similar specs.
Hub Engineer information The Hub Engineer should fill in the metadata below when it available. The Community Representative shouldn't worry about this section, but may be asked to provide help answering some questions.
Deployment information Hub ID: USC PaleoHackWeek Hub Cluster: 2i2c Hub URL: https://paleohack2021.hackathon.2i2c.cloud/ Hub Template: basehub
Actions to deploy