Skip to content

Getting Started Overview

NCShare Compute as a Service (CaaS) and GPU as a Service (GaaS) is a program established and funded by the U.S. National Science Foundation to help researchers and educators to utilize computing systems and services – at no cost to you. NCShare accommodates a wide range of computational needs, from interactive sessions using tools like MATLAB, RStudio, and Jupyter Notebooks, to parallel workflows for large-scale computations.

There are a lot of parts to getting up and going with NCShare. Here's an outline for getting started.

Get an account

First, set up an account so you can access NCShare services. While different services may require different setups, every user will need to create an account through COmanage. Set one up here.

Determine which services you need

Once you've created your account, you can then check the services provided by NCShare. There are various services for researchers and classrooms, and those services may require different initial setups. These services include:

  • Open OnDemand (interactive Jupyter Labs and RStudio applications)
  • NCShare Cluster (direct access to the cluster environment through command line)
  • Containers (encapsulated computing environments in Container Manager useful for shared access in classrooms)
  • Data storage information
  • Data transfer information

Open OnDemand

If you need access to cluster resources for research and would prefer to use interactive applications, such as Jupyter Lab and RStudio, take a look at Open OnDemand. Open OnDemand provides these applications from a web browser, providing you an interactive user interface while using the resources of NCShare.

Access Open OnDemand

Once there, you can see our OnDemand guides for additional information: OnDemand Conda environments, OnDemand JupyterLab, and OnDemand RStudio.

Direct Cluster Access

Some users are comfortable with command line and would prefer to interact directly with NCShare's cluster environment. Those users will need to first enable SSH key authentication from their worksation. Once done, users can follow these instructions for accessing the cluster directly.

Containers

If you need access to a container, you can utilize NCShare's Container Manager service. This allows the community easy access to specialized software packages without installing complex software on your computer.

  • Container reservations last for a semester and provide access via web browsers to packages such as RStudio and JupyterLab hosted in MCNC's datacenter.
  • This is partciularlly useful for courses looking to utlized standardized coding environments for their courses over a semester.

You can read more about containers in the Cluster Software page.

Data Storage

  • Please see the Cluster Computing | Cluster Storage section for more detailed information on storage
  • The NCShare cluster has several shared file systems available for all users of the cluster
  • Each user gets a 50GB share at /hpc/home/<ncshareid> where they can store personal scripts or use for environmental setups
  • Institutional contacts can request a shared user space for projects, up to 100TB. This storage space is appropriate to store data sets that researchers plan to analyze. Projects that need to share data can request a 6-month allocation of space.
  • There is a shared workspace at /work/<ncshare> that users have access to for termporary storage. This is appropriate for data sets under analysis, especially if the analyses generate a lot of output. Users can copy data to this location, run their analyses, then copy the needed results back to their project share. Note, files older than 75 days that are stored here will be purged automatically

Please note that no sensitive data is allowed on NCShare. The above storage options are not intended for long-term storage, and users should regularly move data off of NCShare storage.

Data Transfer

For data transfer, we recommend using secure copy (SCP) or rsync. Users can copy data directly to and from a local device to to their volume in NCShare. Users can also utilize the data transfer options through our Open OnDemand instance. Please see our [Data Transfer] guide for more specific information about using scp or rsync, or our OnDemand Jupyter Lab or OnDemand RStudio guides for information about transferring data using OnDemand.