Marshall University Information Technology in collaboration with researchers and scientists are in the process of building an optimized network scientific research and data transfer funded by NSF CC*DNI award.
The objectives of this project are:
- Improve research data flows and remove any data flow constraints.
- Aggregate network traffic from other network domains.
- Improve high-speed access to online scientific applications and data generated at MU, other institutions, and federal agencies
- Cisco 10 G switches at the endpoint buildings to provide 10G access to researchers’ workstations.
- Cisco switches with 40G capabilities to provide 40G backbone and support the 10g endpoint switches.
- Dedicated and independent 1GE uplink to Internet2 and OARnet
- Implement a science DMZ within the campus network infrastructure to allow the trusted data to travel outside the firewall.
- Find bottlenecks and optimize the network for high-volume transfer of datasets
- Utilize perfSONAR for performance Measurement
- Implement switches supporting SDN
- Support OpenFlow-based architectures, experimentally and in operations.
- Install Data Transfer Node
- Improve Identity Management and Authentication
- Fully implement InCommon over the next year to eighteen months.
- Implement CILogon
- Implement Eduroam
- Access training for key IT staff members for new CI tools (perfSONAR, InCommon, CILogon, Eduroam, etc.
- Continue IPv6 implementation for the entire university computing facilities.
- Selective on-premise deployment to only the services and software that are good fit for local hosting while extending other services to other locations (i.e. Big Data Analytics locally, REDCap remotely or local HPC as a staging resource before going to XSEDE).
- Promote and demonstrate cyber-infrastructure initiatives like XSEDE to the campus research community.
- Develop closer partnerships with other institutions’ IT research computing teams and coordinate hosting meetings and workshops.
- Improve high-speed access to online scientific applications and data generated at MU, other institutions and federal agencies.
Click the images below to expand.