Transcript Efficient management of very large volumes of

Hemera KickOff
October 5th, 2010
Working Group B5
Efficient management of very large
volumes of information for dataintensive applications
Gabriel Antoniu, Jean-Marc Pierson
Challenges
• Tremendous volumes of data (up to
Petabytes), increasing every year
• Cloud infrastructures enforce this trend
• Large span of diverse applications
• Different modalities of data: images, text,
video, raw values
• Distributed, heterogeneous, structured or not,
semantically (en-)riched, confidential
• Stored in DFS or DDB, Cloud storage
services, Warehouses
Aim of the WG
• Explore research issues related to high-level services
for information management (search, mining,
visualisation, processing)
• For large volumes of distributed data
• Taking into account
– security, efficiency and heterogeneity
– applications requirements
– and the execution infrastructure (grids, clouds)
Issues to be addressed
• Low-level:
– Fault-tolerance, caching, transport, security
(encryption, confidentiality), consistency, location
transparency
• Intermediate-level:
– Interoperability among storage systems
– Data indexing
• High-level:
– Data mining, data classification, data assimilation,
knowledge extraction, data visualization
– Metadata management
Communities involved
• Distributed applications
• Distributed systems
– clusters, grids, P2P, clouds
•
•
•
•
Fault-tolerant systems
Databases, data mining
Security
Numerical algorithms
Roadmap
• Identify research teams
– Active in the area of the WG
– With experience in data-intensive
applications on Aladdin-G5K
– And new comers…
• Organize workshops and possibly
schools to share and disseminate
experience and knowledge