Tutorial W03 - Settings things up

Getting started

Login to the cluster

$ ssh –Y USERNAME@euler.ethz.ch

Setup git

Generate ssh key
```
$ ssh-keygen -t ed25519  
```
Print your public key
```
$ cat ./ssh/id_ed25519.pub
```
Copy the public key to Github as done in tutorial 0.
Login
```
$ ssh -T git@github.com
```

Clone repo:

$ git clone git@github.com:Digital-Humans-23/tutorial-w03-YOURACCOUNT.git

Installing conda (Linux / Cluster instructions):

Navigate to home folder
```
$ cd
```

Download and run installer.

$ wget https://repo.anaconda.com/miniconda/Miniconda3-py39_23.1.0-1-Linux-x86_64.sh
$ chmod +x ./Miniconda3-py39_23.1.0-1-Linux-x86_64.sh
$ ./Miniconda3-py39_23.1.0-1-Linux-x86_64.sh

Accept License terms.
Press enter to accept default installation.
Allow installer to initialize Miniconda

Disable activation of conda by default (Optional but recommended)

$ source .bashrc
$ conda config --set auto_activate_base false

We start with conda to set up an environment with the minimum dependencies.

$ conda create --name simpleEnv python=3.9.2
$ conda activate simpleEnv

Navigate to path of repo folder e.g.
```
$ cd <PATH TO repo folder>
```
Install requirements
```
$ pip install -r requirements.txt 
```

Compile on server

Build pycrl.so (python wrapper of pycrl C++ libraries)

$ mkdir build && cd build   
$ env2lmod
$ module load gcc/8.2.0 python/3.9.9 cmake/3.25.0 freeglut/3.0.0 libxrandr/1.5.0  libxinerama/1.1.3 libxi/1.7.6  libxcursor/1.1.14 mesa/17.2.3 eth_proxy
# IMPORTANT: If you use a conda environment, you should add 
# -DPython_EXECUTABLE=<PYTHON INTERPRETER PATH>  (Use absolute paths) e.g.:
# cmake -DPython_EXECUTABLE=/cluster/home/mimora/miniconda3/envs/simpleEnv/bin/python3 -DCMAKE_BUILD_TYPE=Release ../
$ cmake -DPython_EXECUTABLE=/cluster/home/<YOUR_USERNAME>/miniconda3/envs/simpleEnv/bin/python3 -DCMAKE_BUILD_TYPE=Release ../
$ make 
# Return to repo folder
$ cd ..

Run jobs on server

Run job
```
$ sbatch ./jobs/01_simple_job   
```

Monitor job

$ watch -n 1 squeue
# Press Ctrl+C to stop monitoring the job

Check that job was run successfully (It should print Done )
```
$ cat ./jobs/simple_job.out    
```