Getting Started
Accessing the Testbed
Remote Desktop Software
One of the easiest ways to access the NETWA is through the a remote desktop software, such as Remote Desktop Connection, which is a pre-installed softare on Windows. Microsoft also has a version available for MacOs as well, but any software that allows remote desktop access will do. Below are steps to access your own virtual desktop on the testbed:
- Launch Remote Desktop Connection or similar software to connect to your virtual desktop
- In the "computer:" field, enter "ciroh-testbed.uvm.edu"
- Click "connect" and enter your university username and password when prompted
- Congratulations! You've just logged on to your very own NETWA virtual desktop
SSH
Another way to access the testbed is through the Secure Shell Protocol, more comomnly known as SSH. This option is more appropriate for users who are familiar with using the Linux terminal or other command-line interfaces.
- Open a terminal application on your machine (there are built-in terminals on MacOS and Windows, or you can use a third-party app such as PuTTY or MobaXterm)
- Type the command
ssh your_netId@ciroh-testbed.uvm.edu
, using your actual netID - Enter your password when prompted
- Now you should be logged in to the testbed and be able to navigate via the command-line
Data Storage
The /netfiles/ciroh/
directory on the NETWA has over 40 terabytes of disk space that researchers can use to store data - that's equivalent to 40 million megabytes! Needless to say, there's plenty of space for data. If you are using any of the data modules in forecast-workflow
that require a directory in which to store downloaded files, such as nwm_fc.py
or gfs_fc_thredds.py
, please use "/netfiles/ciroh/downloadedData/
" as the download path. These modules will check the download path to see if the data you are requesting is already downloaded before attempting to get it again. Using a common shared path for large downloadable datasets like the NWM and GFS ensurse an effiecint use of disk space and also saves time by not having users redownload existing data! You can specify this download path in the module get_data()
functions with the data_dir
parameter. If you have other private data that doesn't make sense to share, please create your own directroy in /netfiles/ciroh
titled after your netID. For example, "John Smith" would store their persoanl data under /netfiles/ciroh/jsmith
.
Setting up Mamba / Conda
Mamba is a lightweight version of Conda, a popular package manager for a variety of programming languages. It is used on the testbed to set up virtual environments that contain all of the relevant packages and dependencies for a given software repository or workflow. There are a few existing mamba/conda Python environments on the testbed already, such as one that houses all of the packages necessary to run the forecast-workflow repo (more specific instructions for accessing that here). This section will demonstrate how to initalize mamba or conda for a new user and how to see what virtual enviornments are available.
- On the testbed, open a terminal and run the following command:
/usr/local/miniforge3/bin/mamba shell init
to install mamba or '/usr/local/miniforge3/bin/conda init' to install conda or both to install both!- Note: you only have to do this once, not every time you open a new terminal or want to use mamba or conda.
- If the command ran successfully, you should see something like this now at the command line:
(base) [jsmith@ciroh-testbed ~]$
- The
(base)
text indicates that you are in the base environment. To see the list of packages in said enviornment (or any environment you happen to be in), runmamba list
orconda list
. - To see a list of available environments, run
mamba env list
or, if using conda, 'conda info --envs'.- Most relevant environments will be located in
/data/condaShared/envs
. - Note: This is one of the VERY few commands that differ between mamba and conda. Thus, you can mostly use
mamba
andconda
at the command line interchangeably.
- Most relevant environments will be located in
- To activate an environmnet, simply run
mamba activate env_name
replacingenv_name
with the actual name of the environment.