EPOC Resources for Campus Cyberinfrastructure (CC*)

About Campus Cyberinfrastructure (CC*)

The National Science Foundation’s Campus Cyberinfrastructure (CC*) program  invests in coordinated campus-level networking and cyberinfrastructure improvements, innovation, integration, and engineering as driven by specific science, research, or educational requirements or use cases.

Across the five different program areas, EPOC can support proposers in several ways as they work on proposals and after project award. These include:

The EPOC team can be reached at epoc@tacc.utexas.edu.

Slides and links to talks for EPOC’s support of the CC* program are available here.

We strongly recommend that all proposal submitters read the solicitation in detail to independently verify all proposal requirements. General assistance on grant writing should be available at your local institution. General information on standard components of NSF grants (budgets, letters of collaboration, bio sketches, etc.) are available in the NSF Proposal & Award Policies & Procedures Guide (PAPPG). When in doubt about specific grant requirements, we recommend contacting the relevant program officers listed in the solicitation.

Identification and description of research or educational drivers to support the requested/upgraded technology (part of all CC* Areas)

Use cases for science and education are required for all Areas of the CC* program. It is critical to understand the end-to-end drivers that support the technology’s needs in order to be successful in this program. The science drivers should come first, and the technology solutions should follow.

Most successful proposals include the following aspects:

When planning cyberinfrastructure (CI) implementations, the EPOC team strongly recommends building a team to understand and meet their needs:

  • Ensuring that what you want to request/build will meet the needs of your end users. While a particular technical component, such as a DMZ, may be of strong interest to the IT team to build, it must be needed by the science drivers. For example, if the biggest pain point for the researchers is the need for additional compute resources, the proposal should not focus on a request for a DMZ.
  • Keep in mind these proposals, and the work that follows, are meant to be collaborative efforts between researchers, educators, IT administration, and CI implementers. To be successful, it is important that each member have a say in the work that is defined and that the  lines of communication are kept open.
  • Understand that there is no perfect or optimal  implementation, those only exist as theoretical or mathematical concepts in this space. It is critical to define an approach that meets the needs of a set of specific use cases, as opposed to one that is theoretically perfect.
  • Start by listening to understand the user’s needs, as opposed to listening to reply with a solution.  The first step must be to understand requirements fully.
  • We strongly recommend proposers “Talk twice, implement once”. What this means in practice is gathering up use cases, designing infrastructure, estimates of how it will impact the use cases, and then verifying again with the users to make sure that what is envisioned will in fact address the pain points.

In general, we have often seen the people-issues and meeting research needs be overlooked in a rush to technology’s choices.

Finding the relevant use cases at an institution can be challenging for some IT groups. We’ve had luck in doing this by looking at:

EPOC also supports the use of Campus Deep Dives for a site to better understand their internal science drivers. This is a roughly 6 month process that uses a structured conversation around case studies to analyze the CI needs more broadly. Additional information on Campus Deep Dives is available at: https://epoc.global/media/filer_public/02/82/02822c5b-06ab-4743-a319-acda71c7290b/application-deep-dive-description-1.pdf

Letters of Collaboration are supplementary documents that are part of NSF grant proposals to detail the full team and collaborators for a project. The NSF PAPPG guidelines for Letters of Collaboration list a recommended format, but this is NOT required for the CC* program, which gives proposers the opportunity to add details about the nature of the collaboration. Each letter MUST include some form of collaboration – they cannot simply say a project is a good project, there must be interaction at some level included. If you would like a Letter of Collaboration with the EPOC project, guidelines are here: https://epoc.global/proposal-collaborations/

Before you submit the proposal, EPOC can help you use NetSage to identify your users sharing data between institutions. Please contact us at epoc@tacc.utexas.edu.

Complete network solution/architecture along with diagrams (part of Areas 1, 2, 3, and 4) 

One of the principles of the CC* solicitation is to include a full technical solution including diagrams (except for the planning grant category), and information about how it will support modern scientific or education use cases in context. The important components for this activity include:

In addition, the EPOC team can offer advice on preliminary designs and architectures, both during proposal development and upon award. Please contact us at epoc@tacc.utexas.edu.

The Science DMZ is a portion of the network, built at or near an institution’s local network perimeter that enables more efficient data transfers for high-performance research applications through tuning of the equipment, configuration, and security policies, as opposed to general-purpose business systems or “enterprise” efficiencies.

For additional information, or to discuss your proposed architecture, please email epoc@tacc.utexas.edu.

Data mobility, the ability  to predictably and efficiently move scientific data between experimental sources, processing facilities, long term storage, and collaborators, is a common use case that transcends the boundaries of research disciplines. This is generally supported through two components, the system facing Data Transfer Node (DTN) and the user facing Data Portal.

For additional information, please email epoc@tacc.utexas.edu.

A strong Measurement and Monitoring Framework can help implementers know what performance to expect from data transfers and to find bottlenecks that might occur in the network, the disk, or the CPU. The 2020 CC* solicitation  specifically calls out the use of perfSONAR as part of this implementation.

For additional information, please email epoc@tacc.utexas.edu.

Campus CyberInfrastructure plan (part of Areas 1, 2, 3, and 4)

All proposals submitted to the CC* program, with the exception of area (5), must include a Campus CI plan within which the proposed CI improvements are conceived, designed, and implemented in the context of a coherent campus-wide strategy and approach to CI that is integrated horizontally intra-campus and vertically with regional and national CI investments and best practices. This Campus CI plan must be included as a Supplementary Document and is limited to no more than 5 pages.

There is no standard definition of what is included in a Campus CI plan, however, an archive of plans submitted in previous proposals is available at:

https://fasterdata.es.net/campusCIplanning/. In addition, the EPOC team is available to review and comment on draft Campus CI plans, please email epoc@tacc.utexas.edu.

Materials

CARCC Emerging Centers presentation – December 16, 2020

Quilt presentation – December 15, 2020