1
Feburary 8, 2010
DataSpace
1
HP Labs Research Interests • HP Labs have organized its corporate research around 8
major themes that include Information management, Analytics, and Cloud computing– Information Management addresses the need to
govern and integrate multiple sources and formats of data.
– Analytics focuses on turning enormous and rapidly increasing amounts of data into relevant insight.
– Cloud research is focused on delivering an application and computing end-state of Everything-as-a-Service
• HP considers breakthroughs in these areas vital to the industry, and these interests intersect well with those of the DataSpace project
2
HP Labs Team • Meichun Hsu– Director, Intelligent Information Management Lab– Focused research on Live Analytics Platform
• Dejan Milojicic– Senior Research Manager, Strategy and Innovation
Office, HP Labs– Managing Director, Open Cirrus
• Joe Pato– Distinguished Technologist, Systems Security Lab– Collaborating with MIT Decentralized Information Group
3
Live Analytics Platform• Research Goal: Develop breakthroughs to create a large-scale
data-intensive analytics platform that combines massively parallel data management with massively parallel data analytics, to extract insights from a combination of structured, unstructured, static and streaming data.
• Live Analytics Platform can be an experimental vehicle for the Dataspace project – Develop new approaches to integrate data of diverse data
formats– Develop novel scientific analytics that require massive data and
computing scale– Offer a computation environment to perform streaming
analytics over multiple sources of scientific data to facilitate low-latency information integration, monitoring, and scientific discovery
• Live Analytics Platform can be offered in the context of Open Cirrus, HP Labs’s experimental Cloud Platform
4
Open CirrusOpen cloud-computing research testbed
Open Cirrus Goals• Foster new systems research in cloud computing• Catalyze open-source stack and APIs for the cloud
What is Open Cirrus• Infrastructure for research spanning 10 sites of 1000+ cores each• Collaborative community sharing research, software, data sets, and best
practices• Sponsored by HP, Yahoo, Intel
Why are Open Cirrus and DataSpace a match• Federation of Open Cirrus sites matches―DataSpace distributed federated
ecosystem• Collaboration among 10 Open Cirrus sites―interdisciplinary researchers,
applications• Large diverse data sets―mutually beneficial• Global testbed―ideal for experiments on automation of security, privacy, export
rules
http://opencirrus.org
5
Information Accountability
• HP’s Collaboration with MIT Decentralized Information Group– Investigating techniques to encourage responsible
use of information in a world of global, open, and interconnected systems
– Explore the inclusion of flexible security models for provenance in the DataSpace infrastructure
6