Stadler Lab Server SOP

Table of contents

1. Important: Before starting any job
2. Getting access to the server
3. Connecting to the server
4. Setting up Conda
5. Setting up Jupyter Notebook in VS Code

Important: Before starting any job

Check CPU and RAM usage with htop

• p00 has 64 physical CPU cores (not 128 as shown in htop) and 996 GB of RAM. If you need to run a job that requires > 32 CPU cores or > 512 GB of RAM, please post a note in the Slack channel first so other users are aware.

Check disk usage with df -h

• Your home directory is located on the /home partition (/home/Users/<your-netid>, e.g., /home/Users/xw66). However, it is not recommended to store anything in your home directory, including data, code, databases, etc., unless absolutely necessary. This is because the disk array is old and unstable, and data loss could occur at any time. Ideally, your home directory should only store configuration scripts (e.g., .bashrc) and temporary files (e.g., .bash_history).
• Use /dodo for data storage. Move any data in your home directory into /dodo/<your-netid>. If you don’t have access (e.g., it’s not listed in df -h), contact Michael for help.
• Use /data to store datasets that need fast disk I/O (e.g., BLAST databases).

Getting access to the server

1. Fill out this form and select Dr. Stadler as sponsor (https://www.crc.rice.edu/app/ricelogin.php).
2. In "Software Requirements or Comments", write:

   Request access to lab server p00.cs.rice.edu and file system sno.cs.rice.edu:/dodo. 
   

3. Once Dr. Stadler has approved the request, Rice IT will add you as a regular user on the server.

Connecting to the server

Via terminal

• In your terminal, type

  ssh <your-netid>@p00.cs.rice.edu

  For example, ssh xw66@p00.cs.rice.edu
• When prompted, type yes to continue.
• Enter your NetID password. If successful, you will be connected to the server at your home directory.

Via VS Code (recommended)

VS Code can also connect to the server via SSH and provides a user-friendly IDE for managing files, writing code, and running jobs.

• Install VS Code from the official website (https://code.visualstudio.com/).
• (Optional) Link your GitHub account to sync your settings.
• Click the "Extensions" icon on the left, and search for "ssh".

• Install the "Remote - SSH" extension.

• Click the new icon that appears on the left sidebar, called "Remote Explorer". Click the "+" button.

• In the pop-up text box, enter the SSH command

  ssh <your-netid>@p00.cs.rice.edu

  For example, ssh xw66@p00.cs.rice.edu
• Select the default SSH config file. In the pop-up window in the bottom-right corner, click "Connect".
• A new VS Code window will open. Select "Continue" and enter your password.
• In the new window, click the "Remote Explorer" icon on the left, and you should see that you are now connected to the server.

• In the center of the window, click "Open...". In the pop-up window, navigate to the directory you want to open and click "OK".
• Drag-and-drop any files to the file explorer on the left to upload them to the server. Right-click a file to download it to your computer.

Setting up Conda

1. Install conda
Conda is installed with Miniforge. See their installation guide for Unix-like platforms.

2. Set up conda channels

conda config --remove-key channels
conda config --add channels bioconda
conda config --add channels conda-forge
conda config --set channel_priority strict

This will create a .condarc file in your home directory.

3. Create a conda environment for your project
Each project should have its own environment, and you should never use the base environment.

# replace <myenv> with your project name
# add packages needed for your project
conda create -n <myenv> numpy pandas

4. Activate the conda environment
You must activate an environment before using packages installed in it.

conda activate <myenv>

If successful, the name of the environment will be shown in the shell prompt. For example,

xw66@p00:~$ conda activate seqwin
(seqwin) xw66@p00:~$ 

Setting up Jupyter Notebook in VS Code

1. Install VS Code extensions on your local machine
Before connecting to the server, go to the Extensions tab (left sidebar), search for and install the Python and Jupyter extensions.

2. Install VS Code extensions on the server

• After connecting to the server, go to the Extensions tab again. Click the download button in the second dropdown menu labeled "SSH: ...".

• In the pop-up window, select all extensions and click "OK".

3. Create a notebook file

• In the VS Code menu, click File > New File... > Jupyter Notebook.
• In the empty notebook, in the upper-right corner, click Select Kernel > Python Environments, and select one of your conda environments.