Running a job

For this tutorial I'll be creating an empty folder called tutorial in my sharedscratch

The VS Code way
1. Right click on the sharedscratch folder and select New Folder...
2. Name the folder tutorial
Shell
```
$ mkdir sharedscratch/tutorial
```

Let's say I want to run a Python script called test.py on Maxwell, it's just a simple script that will print This is a Maxwell test 100 times

# sharedscratchtutorial/tutorial/test.py

for i in range(100):
    print("This is a Maxwell test")

The VS Code way
1. Create a new file by right clicking on the tutorial folder
2. Name the file test.py
3. Copy and paste the code above to the file
4. Save the file by pressing CTRL+S on Windows or ⌘+ s on Mac
5. You can open a terminal CTRL + ` on Windows or ⌃ + ` on Mac, then run the code by typing python test.py and should be working smoothly for something small like this, but in reality you'll be working with much bigger stuff and its not recommended to run your whole pipeline on a login node, so instead of doing that we will use SLURM
6. Create a new file called slurm.sh in the same directory as your test.py
7. Paste the following blocks of code to the slurm.sh
```
#!/bin/bash
#SBATCH --job-name=tester
#SBATCH -o tester.out
#SBATCH --ntasks=1
#SBATCH --mail-type=begin
#SBATCH --mail-type=end
#SBATCH --mail-user=m.ramdhani.23@abdn.ac.uk

python test.py
```
  The first line is called the shebang it specifify the interpreter to be used to run the code, next lines starting with #SBATCH are the special instructions we'll use to interact with the Slurm Workload Manager
  - --job-name: Name for the job
  - --o: Output log
  - --ntasks: Number of tasks to be executed in parallel
  - --mail-..: Send a mail when the job starts running and when the job is finished to --mail-user email address
8. Once we're set we can run the code from the terminal
```
# Change the directory if you haven't done so
$ cd sharedscratch/tutorial

# Run the script
$ sbatch slurm.sh
```
1. You can check the status of the job by using squeue -u <YOUR_USER_ID>
2. You can also see the log once your job starts running in a file called tester.out
Shell
```
$ cd sharedscratch/tutorial

$ vim test.py
```
esc then i to insert text, ctrl + v to paste the python code, save it by pressing esc then w then q and enter to save file
```
vim slurm.sh
```
Do the same thing above but this time paste the code for slurm.sh and save it
```
$ sbatch slurm.sh
$ squeue -u <YOUR_USERNAME>
```

When running your job you might feel like you want to update your script or stop the already running job to do this you can use scancel <JOB_ID> JOB_ID of your job can be found on the first column when you run squeue -u <YOUR_USERNAME>

More resources: