Lesson plan¶
- Date: 20240827
- Teacher: Richel
- Course: intro to UPPMAX day 1
Learning objectives¶
Here I collected all the learning objectives from the original course:
- Let's step up and have an overview of SNIC
- This is a short introduction in how to reach the calculation nodes
- Wednesday afternoon is wedded to this topic!
- We'll get an overview of UPPMAX and SNIC/NAISS and how a computer cluster works
- Let's recap and go a little deeper into the organization of NAISS/UPPMAX
- Let's step up and have an overview of NAISS
- We'll relate our login session to the specific "area" in the cluster.
- We'll go through platform specific (Mac/Linux/Windows) ways to log in to UPPMAX.
- We'll briefly get an overview of Linux
- How the command line works
- Some text editors
- Things to be aware of
- Let's dig into the most important BASH commands
- We'll do a type-along session
- Get overview of UPPMAX and NAISS
- Work with your files in a LINUX environment
- Learn to use the UPPMAX resources in a good way
- Learn the basics in the Python programming language
- Let's recap and go a little further into the UPPMAX hardware!
- We'll briefly get an overview of kinds of sensitive data
- ... and the Bianca system
I will use these as a starting point.
My goals¶
- Use the cluster guides of the UPPMAX documentation as much as possible:
- No duplicate information
- Ensures UPPMAX documentation is of high quality, as it is actually used in teaching
- Use Linux info from other source, such as https://linuxjourney.com
Method¶
I started with the course content of https://github.com/UPPMAX/uppmax_intro and then shuffled it around.
Starting from the connectivist theory of learning, I decided to do visual things first (among others, file transfer using FileZilla), then using the terminal (among others, file transfer using scp) and only then discuss theory about more theoretical things (such as: NAISS, Bianca).
Working on this course puts requirements on the UPPMAX documentation too, as this is the course materials I will be using.
Progress¶
I've molded the original course content in what it was now. Now for the scheduling: I am going to take the time I needed for all exercises, multiply that times 3, then see how full the schedule is.
| Session | My time (mins, rounded up) |
|---|---|
| Basic use visual tools | . |
| login | 6 + 8 |
| login_remote_desktop | 5 |
| use_remote_desktop | 12 |
| file_transfer_filezilla | 7 |
| Basic use terminal | . |
| login_ssh | 7 |
| use_terminal | 16 |
| file_transfer_scp | 6 |
| Real work | . |
| modules | 9 |
| scripts | 8 |
| job_scheduler | 17 |
| interactive | 7 |
| Rest | . |
| use_ide | - |
| naiss | - |
| clusters | - |
| courses | - |
| Total | 108 |
This fits badly, as I have 300 mins of teaching for (108x3=)324 minutes of exercises! I was right to ignore the misc things! Now, let's break it down in a schedule:
| Session | Time needed (mins, rounded up) | Progress when done (%) |
|---|---|---|
| Basic use visual tools | . | 38 |
| login | 52 | 16 |
| login_remote_desktop | 15 | 21 |
| use_remote_desktop | 36 | 31 |
| file_transfer_filezilla | 21 | 38 |
| Basic use terminal | . | 64 |
| login_ssh | 21 | 44 |
| use_terminal | 48 | 58 |
| file_transfer_scp | 18 | 64 |
| Real work | . | 100 |
| modules | 27 | 72 |
| scripts | 24 | 79 |
| job_scheduler | 51 | 94 |
| interactive | 21 | 100 |
It seems to fit nicely in blocks of 2. Let's see what happens! I've made scp optional, as FileZilla is good enough.
My biggest worry is at the start, when the learners need to log in. I simplified the UPPMAX documentation with some extra movies.
I will use group sizes of 3, due to my last reflection being positive on group sizes of 3.
I don't need to put much effort in a lesson plan here, as (1) I run student-paced, (2) the material is nicely laid out
I posted the Google Form links on the course website already. No bots picked it up: zero responses
From the SUPR NAISS project, I see that there are 23 learners. Using the golden ratio, 0.618033989 x 23 = 14.2 = 15
How to manage the groups? I think the exercise procedure does make sense, let's stick to it. One parameter, however, is group size. I want 8 groups max. Yup, so 3 per group will work.
9 already logged in to Rackham for sure: