vldata common solution for the (very-)large data challenge

13
VLDATA Common solution for the (very-)large data challenge EINFRA-1, focus on topics (4) & (5)

Upload: emele

Post on 22-Mar-2016

47 views

Category:

Documents


0 download

DESCRIPTION

VLDATA Common solution for the (very-)large data challenge. EINFRA-1, focus on topics (4) & (5). Objectives. - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: VLDATA Common solution for the (very-)large data challenge

VLDATACommon solution for the (very-)large data

challenge

EINFRA-1, focus on topics (4) & (5)

Page 2: VLDATA Common solution for the (very-)large data challenge

Objectives An open & generic platform supporting efficient and

cost-effective solutions for large-scale distributed data processing, curation, analysis and publication. Providing standard-base interfaces and interoperable access to various e-Infrastructures

Evolution of existing solutions, advancing state-of-the-art, addressing: openness, extensibility, flexibility, interoperability, scalability, efficiency, productivity, security, cost effectiveness

User Community driven co-design, validated by end users and supporting a new generation of data scientists.

Cooperation among Technology providers, integrating existing technologies, to simplify the connection between Users and Resource providers.

Page 3: VLDATA Common solution for the (very-)large data challenge

(Direct) Impact To the Research Infrastructures Scalability, robustness Participating RIs will operate their Distributed Computing Systems efficiently,

processing their large volume research data, making it available to their end users in a reliable and cost-effective manner that couldn't be achieved before. This may lead to new ways of organizing science activities, leading to significant scientific breakthroughs.

To the end user: scientist/operator Simplicity and interoperability By providing important functional components (e.g., pilots, single interface, etc. )

missing from existing practices, VLDATA platform will make possible the transparent integration of resources, hiding the complexity from use, resulting in the extension of the scale of the resources RIs can utilize.

To funding agencies Cost efficiency By reducing the duplication efforts, maximizing the use of EU-invest on e-

Infrastructures, enlarging the user communities, providing efficient data processing services, providing advanced technology by integrating the state-of-the-art will reduce development cost significantly.

Page 4: VLDATA Common solution for the (very-)large data challenge

Consortium

Page 5: VLDATA Common solution for the (very-)large data challenge

Make IT simple Simplicity: VLDATA provides an abstraction of the different Resources that are all

made accessible the end user via the same interfaces. Transparency: Users are allowed to specify their Workflows/Pipelines with

different levels of abstractions. The platform takes care of the necessary Resource Allocation to fulfill the required specifications.

Extendibility and flexibility: VLDATA provides an API that allows users to extend the provided functionality by developing new or customized components

Reliability: Quality standards and extensive validation in several scientific domains to ensure the readiness-to-use and robustness of VLDATA based solutions

Scalability: Modular implementation allowing horizontal (amount of connected Resources or Users) and vertical (amount of processed Units) scaling to adapt VLDATA to the needs of each particular community or Research Infrastructure project.

Smart and intelligent: building on collected experience and monitoring data, algorithm can look for optimized scheduling/searching strategies, including automated decision making based on usage traces and expectations.

Cost-effective: Building up on existing well-established solutions and incrementally extending and developing to address new challenges with an evolving validated common solution, avoiding unnecessary duplicated efforts.

Page 6: VLDATA Common solution for the (very-)large data challenge

VLDATA

Framework: System Logging, Configuration, Accounting, Monitoring

Adva

nced

M

odul

es: C

ompu

te

& Da

ta M

anag

emen

t

Basi

c M

odul

es:

File

Cata

log,

Res

ourc

e St

atus

, Req

uest

M

anag

emen

t, W

orkl

oad

Man

agem

ent

Use

r In

terf

aces

:Po

rtals,

Com

man

d Lin

e, R

EST,

API

s

Reso

urce

In

terf

aces

:Gr

ids,

Clou

ds, C

lust

ers

Com

putin

g, S

tora

gePu

blic,

Priv

ate,

Vo

lunt

eer

Requ

irem

ents

Qua

lity

& S

ecur

ity

Page 7: VLDATA Common solution for the (very-)large data challenge

Organization

Page 8: VLDATA Common solution for the (very-)large data challenge

8

Development Area (WP 1-5)

Main Partners: AMC, Cardiff, CPPM, CYFRONET, DESY, MTA SZTAKI, UAB, UB, Westminster

Working Cycle: Requirement-Analysis -> Design -> Development -> Integration -> Quality

Control Work Plan: Year 0: Prototype Year 1: Scaling + Integration Year 2: Catalogs + Quality Year 3: Virtualization + Security Year 4: New Challenges + Consolidation

Sustainability Open Source Distribute Data Processing Collaboration: Open DISData

Association 4-year decreasing budget for Development Area

Page 9: VLDATA Common solution for the (very-)large data challenge

Validation Area (WP 6) Each participating RI produces and validates it own

solution using common framework and tools Main Partners

LHCb, Belle II, BES III, Pierre Auger Observatory, EISCAT_3D, Astrophysics, Computational Chemistry, Molecular Structure simulation, Seismology

(TBC) IceCube, COMPASS, NA62, CTA SMEs

Working Cycle: Design -> Integration -> Validation

Transversal activities, sharing experience, training, tools, etc.

Page 10: VLDATA Common solution for the (very-)large data challenge

Exploitation Area (WP 7-9)

Target: sustainability Main partners

CNRS, UvA, EGI, ASCAMM, ETL, Bull From “DIRAC Consortium” towards “Open

Distributed Data Processing Collaboration” Work Packages:

Communication Training Sustainability

Page 11: VLDATA Common solution for the (very-)large data challenge

Outputs & Inputs

This Consortium

WP1

WP2, 3, 4, 5

WP6

WP7, 8, 9

Experts

Resource Providers

Other Projects

Other RIs

Existing Products

Policy Makers

Page 12: VLDATA Common solution for the (very-)large data challenge

12

Working model User community driven co-development

(Rapid Application Development): Open, iterative, incremental and parallel,

requirement-driven development process

Page 13: VLDATA Common solution for the (very-)large data challenge

AbstractThe proposed project aims to produce and validate common solutions to the processing,

curation, analysis and publication of very large scientific data generated by European and world wide scientific Research Infrastructures (RIs). The number of RIs in Europe and beyond expected to collect yearly multi Petabyte data samples increases exponentially and they will soon be reaching the Exa scale. Existing solutions must be evolved in order to cope with large scale distributed data processing. The VLDATA platform will provide standard based interoperable access to various types of resources: Grid, Cloud, Volunteer, HPC, etc. (funded with different models: capex or opex, and coming from the public or private sector), and software tools/services running on top to support global data science. Various RIs from different scientific domains, from physics to life sciences or to chemistry will validate VLDATA platform by implementing solutions for their concrete use cases, achieving at the same time a significant optimization in the efficiency and cost, and ensuring that no aspects of the challenge will be ignored. The complete life cycle of the data will be addressed, as well as interoperability between different scientific domains and e Infrastructures.

The project gathers experts with complementary backgrounds from the Technology and the Resource Provider worlds that, collaborating with other relevant external experts and those from the participating RIs, will: (1) analyse the requirements for each of the RIs, (2) provide the VLDATA generic platform, starting from current solutions and following an incremental iterative development model, (3) design, prototype and implements the Distributed Computing Systems for each of the projects participating RIs, and (4) make the resulting VLDATA platform available to other RIs with similar needs. To reach other RIs, VLDATA will promote standard interfaces and tools, define appropriate quality assurance mechanisms and provide dissemination and training events, aiming to be sustained in the long run by contributions from new RIs benefitting from the VLDATA platform.