Technical Storage Architectures

Attendees: HEP operations, NSF DataNet and BIGDATA, Digital Library community.

Location: As a satellite of a visible national meeting involved in large scale storage, such as an ACM data management conference (SIGMOD,VLDB,XLDB), ACM Supercomputing or the IEEE Mass Storage Systems conference (MSS).

Purpose: his workshop will bring together stakeholders that are (or will be) involved in the operation of real sites that must store large data volumes to exchange information and experience on running scalable ´┐╝storage systems. Each of the invited communities will bring a different perspective: the HEP operators on the unique needs of the HEP software, the BIGDATA community with a general CS perspective, and the ´┐╝DataNet and digital library communities with expertise in long term preservation.

Inputs: Design documents for commonly used storage architectures, experience reports from DataNet projects, industry experience with BIGDATA technologies, latest developments in university digital libraries and research computing centers.

Outcomes: Collect best practices in designing, procuring, and operating large scale storage systems; interactions between technology and capability, such as what software systems are appropriate for what workloads; and experiences with designing and enforcing organizational policies.