Session IV: Tiering in and Peering out: Approaches to Data-Intensive Science

Today’s state of the practice in data-intensive science requires us to take two different perspectives – one within the facility, and one across facilities. Looking inwards within a computing facility (CF) we focus on tiered storage and storage representations to improve access and manage data for analytics-driven discovery at scale. Looking outwards we need to create peering relationships with Observational Facilities (OF) that are sources of large-scale data. We highlight the considerations at play in such peered connections across CFs and OFs. The seamless overlay across and within facilities is a future state that demands simplified computing and data access models that connect the data from the edge, through the fog of middleware, to scalable computing and analytics resources in the supercomputing cloud.

Download Presentation

Location: Grand Ballroom C Date: March 28, 2019 Time: 8:00 am - 8:20 am Arjun Shankar, ORNL