Cloud Connectivity for Data Intensive Science: Considerations and Lessons Learned
Time 10/17/18 02:00PM-02:10PM
As data production in scientific research environments increases, it is becoming increasingly difficult for scientists to share those data effectively due to the sheer volume of data for individual research projects. Additionally, the data density is costing research organizations tens of millions of dollars to support the storage and analysis of the data within their own datacenters. As a result, interest in utilizing public cloud environments for large data storage and sharing to enable collaborative analyses has grown and many organizations are exploring the potential of these resources. However, many of them do not have the necessary connectivity to cloud resources to enable the movement of data to and from clouds, and often bottleneck data transfers through security borders that are designed for enterprise traffic patterns, and not for large data movement. In this brief talk, we will discuss various methodologies for connecting to public clouds at high speed (including science DMZ strategies), as well as report out on lessons learned from various types of organizations (government, industry) who have attempted to enable cloud connectivity. The goal of this presentation will be to add to the group discussions that will provide input for how Internet2 designs its future infrastructure to support the needs of their customers.
Speaker Ari Berman BioTeam, Inc.
Primary track Advanced Networking