Progress of "Plasma and Fusion Cloud" Research Data Ecosystem Orienting "Open Science"

Summary: "Plasma and Fusion Cloud" data ecosystem has been technically verified on some issues. Enormous amount of diagnostic data inevitably require a HPC platform nearby, not only for LHD/SNET but also for ITER, JT-60SA, and other experimental devices. To make the "cloud" data ecosystem may help to exchange the interdisciplinary intelligence not only between experiments and theoretical model calculations, but also with other fields. Cooperation with other community are very essential for "Open Science". 1. Performance evaluation have been made using the HPC supercomputer ^Raijin and the LHD primary data storage. - Directly connected by 100 Gbps Ethernet link. - Throughput improvement is still ongoing. Multiple parallel streams might improve the speed. 2. 1.22 million DOI issued for each diagnostic data object - LHD have roughly 40 million data objects by "data-name  shot-no." basis - Registration fee and the number of years to complete registrations are the problems! - Other fields registering many data DOIs have long history of international data exchanges, cf. geo-science. 3. Open Science – How far as open as possible and as closed as necessary? - LHD has decided open everything: data, codes, proposals, summaries, logs, etc. - Cloud technology and its storage is a powerful supporter for scientific Open Data. - Data license agreements are needed with other SNET institutions, e.g., consensus for common license? 4. Standardization in fusion community - Standardization for data platform; UDA and IMAS compatibility checking and implementation are ongoing.