The scientific world is in a period of opening and sharing of the data it produces. The management of these data is a main concern for the researchers and for the people in charge of laboratories computing infrastructure. They have to cope with a huge amount of informations that have to be preserved for short or long term period.
With its combined experience in high performance computing and long term archiving and its computing infrastructures of excellence, the CINES helps the data producers and managers to handle with their archiving problematics.
The main purpose is to assure the data accessibility, integrity, readability and comprehension as long as it’s needed. The archiving level of demand is adapted according to this duration.
From calculation on Occigen to archiving
High performance computing generates large amounts of data. Some of them deserve to be preserved, for a variety of reasons:
– Either they have required many hours of calculation, which represents a significant financial cost
– Either they have to be preserved for comparison and validation (non-regressive tests)
– Or for both reasons
There are back-up solutions, but they aren’t permanent and most of the time, they last only for the calculation project duration. It’s necessary to lean on an archives service to be sure that the results are preserved when it’s required. That’s why some of these results are transferred to the CINES digital archives service.
So, the processing chain is complete within the CINES, from the data producer to the preservation and communication. This short circuit allows a greater speed between the creation step and the archiving step. It also allows saving because the data don’t have to be temporary stored in different places of the processing chain.
EUDAT (European Data for e-science) is a European project.
Its goal is to set up a collaborative infrastructure providing to the European scientific community a way to manage the data cycle of life.
Started in 2011, EUDAT consortium has been extended in the context of Horizon 2020. Coming from 15 different countries, 35 partners with multiple skills are gathered: computing centers, data centers, libraries, scientific communities, data scientists. CINES is the only French member who is both a data center and a computing center. It is also the only French founding member of the distributed European infrastructure. Some of the services CINES proposes are based on EUDAT tools: “B2SAFE“, “B2STAGE” ou encore “B2HANDLE“.
For more information on : EUDAT
the CINES – HUMA-NUM collaboration
Huma-Num is a “major research infrastructure” (in French, “très grande infrastructure de recherché” or “TGIR”).