Workplan

Full Name

CODATA GOSC Technology Infrastructure Working Group

Short Name

GOSC TI WG

Sign-Off Date

October 2021

Mission Statement

The GOSC TI WG’s vision is to enable technical interoperation between and among digital research infrastructures worldwide in order to support the needs of science internationally, both individual scientists and collaborations of scientists from around the world.

The GOSC TI WG aims to 

1. facilitate interconnection and technical interoperation among, and access to,

2. digital research Infrastructures, from national, regional, community and international levels,

3. by building on and improving existing federation capabilities and interoperation frameworks,

4. in support of international research collaborations and Open Science.

The WG intends that the federation capabilities and interoperation frameworks that result from its work can be adopted by other initiatives to extend the principles of open access to other sectors including public administration and small- and medium-sized enterprises, however this is out of scope for the current effort.

The success of this WG, and GOSC overall, can only be measured by the success of the science it supports.

Objectives

There are numerous important topics for e-infrastructure cooperation, alignment and interoperability, including but not limited to: mechanisms for global federation and cooperation of computing resources, interoperation, capability-based security and AAI model, etc. 

The main objectives of the GOSC TI WG are as follows:

Objective 1: Identify and engage with digital research Infrastructures from around the world.

1.1 Identify and establish contact with digital research infrastructures from around the world.  These are the “participating infrastructures”.

1.2 Engage with Participating Infrastructure contacts and leaders to share current experience and best practice, identify common development areas, work towards cooperation, alignment and interoperability. 

Objective 2: Develop Common Frameworks and Guiding Principles for Interoperation Between Participating Infrastructures

2.1 Collect/assemble existing frameworks/templates, and existing guiding principles, for distributed interoperation that have been created.

2.2 Review features of collected frameworks to identify commonalities, as well as to categorize frameworks into different groups (e.g. AAI federation, vs. interoperation of high-throughput computing infrastructures, vs. interoperation of Kubernetes clusters, vs. federated search/interoperation of data repositories)

2.3 Gather similar technical interoperation frameworks and create “templates” or “recipes” that might be used to allow two unconnected participating infrastructures to enable technical interoperation in the area described (e.g. AAI, federated search) 

2.4 Confirm interest of Science Use Case Study Groups in the frameworks being developed, identify missing capabilities, prioritize based on the needs of science. 

Frameworks should address co-service operation and delivery, and user community support models that can work across country/region/continents resources (data and service). Diversity of architectures among the frameworks will be encouraged as it will create more value for the community in accessing all possible technological implementations.

Objective 3 Test the Templates/Recipes to refine the GOSC Frameworks 

Establish testbeds to verify the concepts and approaches defined by the common GOSC frameworks.  Different testbeds might be set up for each agreed framework area (e.g. OpenStack, AAI federation, etc.) Ideally, to establish connections among the global e-Infrastructure, for example to connect China (CSTCloud) and Europe (EGI), South Africa and Europe, China and South Africa, etc. 

Testbeds should include:

• Community OSC to Community OSC interoperation

  • Access to compute resources -- federated data processing, different virtualization/containerization approaches, heterogeneous hardware types

  • Access to data, and data sharing protocols (making interoperable)

  • Data transfer (policies)

  • Enabling complementarity of infrastructure, e.g. workflows requiring different types of infrastructure can be orchestrated across those infrastructures, rather than requiring every infrastructure to have every kind of resource.

• Remote access to Community OSCs

• Authentication/Authorization/Identity Management (AAI)

• Aligning PID policies and infrastructure (check with other WGs)

Report on each test, improvements needed to the “template”, next steps for future testing.  As templates are fully tested they should be formalized as interoperation frameworks, specifying required initial conditions, steps to take, measures of success.

Objective 4 Collaborate with International Science Use Cases to demonstrate the value of GOSC infrastructure 

To work with international science communities and identify concrete use cases to showcase the usage of GOSC infrastructure. For example, demonstrations on cross-continent data movement; data processing on the GOSC federation from different international e-Infrastructures; supporting global research collaboration that cannot be conducted before without a GOSC infrastructure, etc.  

<Participating infrastructures coordinate with science use cases to plan follow on exploitation and implementation activities, including funding efforts -- share information of funding opportunities, work towards concrete project in order to turn the group discussion into real development >      

Description of Work

Task TI-T1 [Months 1-6]: Identify different frameworks and sets of principles for interoperation (reference frameworks - RFs) that new infrastructures can consider and adopt to enable future interoperation.

• TI-T1.1: AAI access RF

• TI-T1.2: Data access RF

• TI-T1.3: Service discovery RF

• TI-T1.4: Remote compute authorization RF (may include several versions, e.g HTCondor, OpenStack, Kubernetes, etc.)

• TI-T1.5:, Storage authorization RF

• TI-T1.6: Remote Compute Execution RF (see versions from TI-T1.4)

• TI-T1.7: Security Cooperation RF

<other areas to be defined based on input from Science Use Case Study Groups>

Established OSCs [e.g. EGI, ARDC, CSTCloud, OSG, etc.]: develop/draft RFs or identify appropriate existing frameworks in each area

• E.g. AARC Blueprint for AAI

• Existing links between OSCs (e.g., between EGI and CSTCloud) can provide starting points for various RFs.

New OSCs [Africa Arab Science and Technology Cloud (AAScTCloud), Malaysian Open Science Cloud [MOSC], African Open Science Platform [AOSP]: to participate in this effort along with more established infrastructures, to validate clarity and applicability of each RF; possibly identify other needed RFs.

Task TI-T2 [Months 7-12]: Define interoperability model, without focussing on technology.  What are the key services that need to exist, that must be provided to users, and that must interoperate between different infrastructures? 

Capture these points as white paper(s) to share with the community and work toward agreed models that might become standards.  In each conceptual area, present best practices of how services, etc. are being delivered by different infrastructures.  White Paper(s) will explore/explain issues related to centralized vs. decentralized architectures (with a preference for the latter, but identifying any central functions that might be useful -- for example a master registry of OSCs participating in the GOSC, along with contact and technical details).

Task TI-T3 [Months 13-18]: Establish technical test beds to test specific interoperation between infrastructures, ideally using different combinations of platforms.  Implement conceptual models, then test interoperation, proving that the model works.  Expand testing from individual pairs of infrastructures to broader interoperability (many-to-many)

• Established OSCs [e.g. EGI, ARDC, CSTCloud, etc.]: Agree to support interoperability tests with new OSCs

• New OSCs: commit to follow/implement specific Reference Frameworks and to work with one or more established OSCs to conduct tests.

• All participants: document test results, update Reference Frameworks and share with TI-WG as basis for subsequent tests.  

• Once Reference Frameworks have been fully tested, they can be published.

• Once interoperability among specific OSCs has been established in a given area (e.g. data sharing, federated AAI, remote compute execution), specific science use cases can be considered.

• Identify gaps in functionality and features between what has been achieved in the testbeds and the requirements defined by the Case Studies.  

  • Some gaps may require implementation of more advanced features than were originally contemplated in the Reference Framework, but which can be delivered based on the technology.  These gaps should be addressed through further rounds of “template” development and testing (return to TI-T2).  

  • Other gaps may require new development, integration or other innovation and will be considered in Task TI-T5.

Task TI-T4 [Months 19-24]: Expand from test beds to application tests (supporting international use cases) -- scaling up and out to support real science use cases.

• Match Requirements for GOSC capabilities (collected from Case Study Groups) against interoperability achieved in Task TI-T3.  

• Specific requirements could include identification of required data sets, access protocols, scale and authorization required for  compute, metadata required and interconnectivity with non-research institutions.

• Specific requirements may point to the need for harmonized API leveraging global and domain standards (in coordination with the data interoperability WG)

Possible use cases could be expanded beyond defined Case Studies, for example to extreme weather in East Africa with cooperation with IGAD, regional intergovernmental organization.

Task TI-T5 [Months 19-24]: Trial and test innovative technologies identified as potentially responding to unmet needs of the science community, for example:

• evolve and design cyber security standards for research computing workloads

• DPU

• Cloud Vulnerability Testing

• Multi-cloud (integrate commercial clouds)

• Metadata tagging for data discovery

TI-D1 [Month 6]: Draft Interoperability Reference Frameworks (RFs) in different areas of interoperability

• TI-D1.1: AAI access RF

• TI-D1.2: Data access RF

• TI-D1.3: Service discovery RF

• TI-D1.4: Remote compute authorization RF (may include several versions, e.g HTCondor, OpenStack, Kubernetes, etc.)

• TI-D1.5: Storage authorization RF

• TI-D1.6: Remote Compute Execution RF (see versions from TI-D1.4)

• TI-D1.7: Security Cooperation RF

• <other areas to be defined>

TI-D2 [Month 12]: White Papers that define interoperability model, without focussing on technology.  What are the key services that need to exist, that must be provided to users, and that must interoperate between different infrastructures? Are there any clear advantages or disadvantages of different models that can be shared?

TI-D3 [Month 18]: Reports on completed testing efforts:

• Final Interoperability Reference Frameworks in each area

• Identification of OSCs that have completed tests of the final framework and are prepared to accept offers to interoperate, either in general or in the context of specific (mutually beneficial) projects.

TI-D4 [Month 24]: Report on successful implementation of test cases from International Science Case Studies.  Recommendations for next steps

Milestones

Each Deliverable (described above) represents its own Milestone.  Additional Milestones (not linked to Deliverables) may be defined.

Objectives

Contributions the to GOSC overall mission 

Targeted Groups

Dissemination Method/Channel

Raise fund for sustaining GOSC development

Sustainability

Funding agencies

Government policy makers

Publications

Alignment, interoperability

Standard body

White papers

Science Communities

Regular meetings

Monthly

Group mailing list

GOSC-WG-Infrastructure@lists.codata.org 

Contacts of Co-chairs

Mark Dietrich, mark.dietrich@egi.eu EGI Foundation

Jianhui LI, lijh@cnic.cn CNIC

Happy Sithole, hsithole@csir.co.za,  CSIR

Contacts of Secretary

Yin Chen, yin.chen@egi.eu,EGI Foundation

Zuliang GUO, zlguo@cnic.cn, CNIC

Confluence Workspace

https://codata-gosc.atlassian.net/wiki/spaces/GOSC/pages/41353298  

/Workplan

/Meetings/yyyy-mm-dd

/Deliverables

/Disseminations/Events/xxxx

/Disseminations/publication and presentations

Name

Organisation

Region/Country

Role

Jianhui Li

Chinese Network Information Center (CNIC), Chinese Academy of Science

China

Co-Chair

Mark Dietrich

EGI Foundation

Europe

Co-Chair

Happy Sithole 

South African National Integrated Cyber Infrastructure System (NICIS)

South Africa

Co-Chair

Yin Chen

EGI Foundation

Europe

Secretary

Zuliang Guo

Chinese Network Information Center (CNIC), Chinese Academy of Science

China

Secretary

Richard Dennis

Royal Danish Library / Copenhagen University Library 

Denmark

Member

Carmel Walsh

Australian Research Data Commons

Australia

Member

Brian Corrie

SFU/iReceptor

Canada

Member

Essam ALHanash

Alexandria University

Africa

Member

Hrachta Astsatryan

The National Academy of Sciences of Armenia

Armenia

Member

Hadia Yaseen

Arab States Research and Education Network

Jordan

Member

Natalia Manola

OPENAIRE/ University of Athens

Europe

Member

Susan Fairley

Global Alliance for Genomics & Health

Europe

Member

Washington Segundo

LA Referencia, RedClara and Open Science Platform

Latin America

Member

Paul Millar

DESY

Europe

Member

Guido Aben

AARNet

Australia

Member

Hussein Sherief

AASCTC

Africa

Member

Lautaro Matas

LA Referencia, RedClara and Open Science Platform

Latin America

Member

Ahmad Mahdavi

Iran

Member

Chifundo Kanjala

Africa

Member

Milan Ojsteršek

University of Maribor

Europe

Member

Maria Stoica

University of Colorado Boulder

US

Member

Hugh Shanahan

Royal Holloway University of London

Europe

Member

Xinan Yue

Institute of Geology and Geophysics , Chinese Academy of Sciences

China

Member

Chenzhou Cui

National Astronomical Observatories, Chinese Academy of Sciences

China

Member

Zhixiong Zhang

National Science Library, CAS

China

Member

Gang Chen

Institute of High Energy Physics, CAS

China

Member

Ziming Zou

National Space Science Center, CAS

China

Member

Yangang Wang

CNIC,CAS

China

Member

Yanbiao Li

CNIC,CAS

China

Member

Zhenyu Li

Institute of Computing Technology, CAS

China

Member

Jimin Wang

Peking University

China

Member

Jeonghoon LEE

Korea

Member

Hans Pfeiffenberger

CODATA IDPC

Europe

Member

Ikopi Molbeko Gabril Marcel Meieko

Member

Gareth O’Neill

Europe

Member

Pascal Heus

US

Member

Joseph Wafula

Africa

Member

Matthews Mtunmbuka

Africa

Member

Suhaimi Napis

MOSP

Malaysia

Member

Liew Chee Sun

MOSP

Malaysia

Member

Anwar Vahed

CSIR

South Africa

Member

Luis Huerter

Asesoria Consultoria & Cobranzas

Chile

Member

Daniel Berrios

NASA

US

Member

Jean Baptiste Fankam

Cameroon

Member

Keith Russeil

ARDC

Australia

Member

Sa-kwang Song

Research Data Sharing Center, Korea Institute of Science and Technology Information(KISTI)

Korea

Member

Lili Zhang

CNIC,CAS

China

Member

Frank Wurthwein

Interim Director of the San Diego Supercomputer Center

US

Member

Miron Livny

University of Wisconsin-Madison

US

Member

Paulo Manghi

OPENAIRE

Europe

Member

Chenglai Wu

Institute of Atmospheric Physics , CAS 

China

Member

Organisations

Type of Resources

EGI

Cloud

CNIC

Cloud

CSIR/CHPC

South Africa e-Infrastructure

SFU/iReceptor

Scientific Platform (http://www.ireceptor.org )

OSGC

e-Infra