03 Call Digest

[jules32]

Hi Everyone!

Thanks to NASA Openscapes Mentors Matt Fisher from NSIDC and Alexis Hunzinger from GES DISC for teaching this week! 

Please continue experimenting in the Hub, and remember Andy Teucher’s notebook about How to store data in the cloud and delete your intermediate and test files. And share your progress in this GitHub Discussion: What's happening on the NASA Openscapes Hub!? Thanks to Lucas Barbedo (Liu-Zhang team) for sharing about using PACE data (comment + thread). In our next Cohort Call we’ll discuss coding strategies and open communities next time to continue building our skillsets for the Cloud. Below is a light digest of Call 3.

Have a great week!

Julie, Erin, Andy, Liz, and the very awesome NASA Openscapes Mentors

Digest: Cohort Call 03 [ 2024-nasa-champions ]

Openscapes_CohortCalls [ 2024-nasa-champions ] Google folder - contains agendas, recordings, pathways 

https://openscapes.github.io/2024-nasa-champions - cohort webpage

 

Goals: We’ll discuss team culture and data strategies for future us in the cloud.

Task: Have a Seaside Chat and work on your pathway (more details in our agenda)

Continue your Pathway - Trailhead & shifting to “Next Steps,” think about team culture & data strategies. Our next Call 4 is open communities & coding strategies in the Cloud

Discuss Team Culture with your groups 

Coworking (optional): May 9. 10-11:30 PT 

Please ask any questions during Coworking or on Slack. 

A chance to work on your own things socially & ask questions/ screenshare. We share what we’re going to work on, and then work quietly and then check in at the end as well. We also make breakout rooms for Q&A if folks want to screenshare and talk things out.

 

Slide Decks:

Psychological safety (slides) 

Data strategies for Future Us (slides)

Data strategies in the JupyterHub (tutorial) 

 

A few lines from shared notes in the Agenda doc:

Being ok making a mistake in front of my team and showing that mistakes are accepted; There’s a phrase around - i’m going to be trip over my words (that’s not quite it) to note that I’m learning. 

 Explicitly pausing periodically to give people opportunity to talk or ask questions. +1 +1

And then holding uncomfortable silence to give people time

Showing work in progress also is super helpful to create a culture where the expectation isn’t that work is only shared when it’s polished. +1+1

I use the phrase “rainbows and lollipops situation” to try to characterize that the fact that we can never have perfection we just have to find ways to incrementally get closer.

There are a few bosses/professors that seem to always “micro-encourage” any thoughts and questions, from students or folks objectively below them. Things like, “that’s really great question”, “yes; let’s talk about this”, “I’m glad you brought that up.” Sometimes these encourage routine questions/issues, but sometimes they’re crazy ones. The key is consistency; so that dissenting views or risky ideas also feel welcome. +1  +1

Becoming wild and setting a low expectation first! helped others to feel more comfortable being themselves. +1 +1

Eli: “hack” team energy. E.g. start day with a stretch. “Everyone share a picture of your dog/cat/kid!” Triggers you to lighten up, creates happy feelings. Makes the work easier. You’ve increased the trust level of the team to go up by having a genuine interaction or something like that

if you need to use cloud-stored data AND locally-stored data, technically speaking, you can bring that local data in the cloud and do your compute there. Lots of ways. This is just to get you thinking about what would work best for your use cases

These are aspirational goals that have not made it fully into our work. Part of the issue is that those of us on our team are low on the food chain that lack the oversight needed to push these ideas into prime time. However, we are individually working toward implementation so I guess in the end well get there if nothing else but a ground up approach.

+1 +1 +1 +1 + 1

 I would like to learn more about working with confidential data in the cloud +1

We often “over-produce” stuff because storage is cheap and access is easy on local machines. How do we learn what is really important and what we can “toss”?  +1 +1

Wonder how to manage keeping intermediate files in the cloud when use may actually be short term? +1+1

Can we use this workflow to store data in a bucket directly and decrease the data storage cost for the cloud?

that's the idea! S3 is faster and cheaper.