WP3 - Interoperation and coordination of eInfrastructures

Wiki pages

Objectives

  • Support the stability of existing and emerging Regional Operation Centres (ROCs) so as to ensure
  • interoperability of DCIs, with focus on Grids.
  • Maintain a set of guidelines for standard grid interfacing, customised for the type of region.
  • Investigate the emerging cloud solutions and propose the relevant interoperation approaches.
  • Analyse the existing HPC interoperation approaches and propose potential solutions.

 

Description and Tasks

Task 3.1 Interoperations support - Leader: GRNET


The task will deal with the support of existing and emerging Regional Operation Centres (ROCs) so as to ensure interoperability of Grid solutions between Europe and the rest of the world. Support will be provided by WP3 for Regional Operations Centre in terms of functionality requirements and structure, while the daily operational effort will be carried out by the ROCs themselves. The functionality will encompass authentication and authorisation, monitoring, user and operational support, management of Service Level Agreements, helpdesks, etc. The functionality and structure will be customised for the type of region, based on CHAIN-REDS guidelines which will be specified through a dedicated deliverable and will be made available through a wiki. Specific attention will be paid to emerging regions such as Africa and Arab Countries.

 

Task 3.2 Promoting and fostering the adoption of standards in R&E eInfrastructures - Leader: INFN


CHAIN-REDS will promote concrete actions towards Middleware interoperability using the existing liaison with middleware providers in the different regions of the world addressed by the project (i.e. EMI in Europe, Garuda in India, CNGrid in China, OurGrid in Latin America, etc.) and also taking active participation in the international fora such as OGF. The task will also explore solutions based on the use of standards for the interface of applications (e.g. SAGA). The activity will include regular meetings with the middleware developers in order to define agreed steps toward the adoption of commonly accepted standards. These technical meetings will be possibly co-located with Workshops and Conferences in order to maximise the participation of interested people and minimise the travel costs.

 

Task 3.3 Clouds for R&E - Leader: CIEMAT


Cloud as an emerging paradigm will be considered in this task through information exchange and coordination of effort between Europe and other world regions. The European solutions e.g. WNodes (INFN), CLEVER (University of Messina), StratusLab and Oceanos (GRNET), OpenNebula (CIEMAT) will be explored/compared,
in coordination with other regions of the world and specifically with developments in Latin America, India and China. A specific cloud technical survey will be carried out, which will summarise technical cloud solutions available in Europe and the world, specifically focusing on federated clouds for Research and Education. A set of recommendations regarding compatibility will be produced and a group of experts from the emerging scenario selected and invited to participate to Technical tables in the CHAIN-REDS area of interest. The goal of these Technical tables will be to actively promote guidance towards interoperation and “federability” between clouds initiatives outside Europe and the European actors.

 

Task 3.4 HPC landscaping - Leader: GRNET


This task will involve the data exchange and studies on HPC state of the art around the world, in terms of resource sharing and common operations within regions. Based on this, a set of best practices and examples for HPC interoperations and sharing guidelines will be produced. The HPC centres operation across the CHAIN-REDS regions will be surveyed, and comparative studies will be made. The task will investigate also the growing uptake of GPU (Graphics Processing Units) based HPC as intermediate and/or complementary resources with better performance/price and performance/power ratios. The results of this activity will be input to the deliverable D3.4 together with the High Throughput Computing delivered by Grid infrastructures and the on-demand Clouds. The task will show how and if these can be merged in a single stop shopping infrastructure for scientific users.