# RIS Getting Started Workshop

- Ayush Chaturvedi ([e-mail]) HPC Research Facilitator III

## Resources:

- WashU [VPN](https://it.wustl.edu/items/connect/)
- RIS [HelpDesk](https://washu.atlassian.net/servicedesk/customer/portal/2/group/6/create/53)
- RIS [Documenation](https://docs.ris.wustl.edu)
- RIS [Virtual Office Hours](https://ris.wustl.edu/support)
- RIS Open-OnDemand [Portal](https://ood.ris.wustl.edu) [VPN Not Required]
- [Slides](https://wustl.box.com/v/risohidsworkshop)
- Recording [Day 1](https://wustl.zoom.us/rec/share/Gi6N5ao96DQU7dkep3NG8wz9KS9h8NKcji9awyOL-P_M_02I2hBSPHNsKS_Ykic.mpZeFgomVrQfOGru) [Day 2](https://wustl.zoom.us/rec/share/ZB1KuYnFuOsT0891e4u8dE8D1JL--OxJFpr4Jy9KE4LpfV49EmUG5t_d8N2cGuPe.1A-8g6YQfyqO4TQM)
## Day 1

- Connecting to RIS

ssh [e-mail]

> Can also ssh to compute1-client-N.ris.wustl.edu where N can be [1-5]. VPN required IFF not on WashU Campus/Network.

## Day 2

### Jupyter through command line

export LSF_DOCKER_PORTS='8100:8888'
bsub  -Is -n 1 -G compute-workshop -q workshop-interactive -a 'docker(rapidsai/rapidsai:0.18-cuda11.0-runtime-ubuntu20.04-py3.8)' -R 'select[port8100]' /bin/bash
jupyter lab

### Custom NoVNC Session

Documentation also [here.](https://docs.ris.wustl.edu/doc/compute/recipes/tools/local-vnc-viewer.html?highlight=novnc)

**Pre-requisites**

1. Have an SSH key pair. See instructions [here.](https://docs.ris.wustl.edu/doc/compute/recipes/ssh-keypair.html?highlight=ssh)
3. On Mac/Linux PC/Laptop.
4. Connected to WashU Network/VPN.
5. Download [Launchdesktop.zip](https://bit.ly/3TWFNMB)
6. Unzip

#### Install Turbo VNC:

On MacOS:
1. Open Terminal App

brew install --cask turbovnc-viewer 

2. And expect:

brew install expect 

3. update path

echo 'export PATH="/opt/TurboVNC/bin:$PATH"' >> ~/.zshrc

### Batch Jobs

Interactive Jobs

bsub -Is -n 1 -G compute-workshop -q workshop-interactive -a 'docker(ubuntu)' /bin/bash 

Batch Jobs

bsub  -n 1 -G compute-workshop -q workshop -a 'docker(ubuntu)' /bin/sleep 10

Requesting GPUs

bsub –G compute-workshop -q workshops -R 'gpuhost' -gpu "num=1:gmodel=TeslaV100_SXM2_32GB" -a 'docker(ubuntu)' /bin/sleep 10

## OHIDS resources

General Tier:

bsub -n 1 -q general -G compute-ohids -a 'docker(ubuntu)' /bin/sleep 10

Subscription:

bsub -n 1 -q subscription -G compute-ohids-t2 -sla ohids-t2 -a 'docker(ubuntu)' /bin/sleep 10

Condo:

bsub -q ohids -G compute-ohids -R 'gpuhost' -a 'docker(alpine)' /bin/sleep 60

> **Note:** OHIDS subscription tier may be discontinued in near future please use condo tier.

---

## Extra

### Creating Anaconda Environment

cd /storage1/fs1/workshops/Active

mkdir -p wustlkey/conda_env
mkdir -p wustlkey/conda_pkg

export LSF_DOCKER_VOLUME='/storage1/fs1/workshops/Active/wustlkey:/storage1/fs1/workshops/Active/wustlkey'

export PATH="$PATH:/opt/conda/bin"

export CONDA_ENVS_DIRS="/storage1/fs1/workshop/Active/wustlkey/conda_env" 

export CONDA_PKGS_DIRS="/storage1/fs1 workshop/Active/wustlkey/conda_pkg" 

bsub -Is -n 1 -G compute-workshop -q workshop-interactive -a 'docker(continuumio/anaconda3:2021.11)' /bin/bash

conda info
conda create  -n env_name