GridPP2: Network Sector Planning

1. GridPP2: Network Sector Planning Robin Tasker CCLRC, Daresbury Laboratory 5 March 2004

2. Network Sector : Objectives • To provide PPNCG support, participate in EGEE oriented • network monitoring service developments and deployment • within the UK, and development of diagnostic engines for • Grid operations. • 2. To exploit the UKLIGHT infrastructure to utilise switched • dedicated circuits between major centres including UK • Tier1 and 2, CERN and FNAL. • 3. To maintain the strategic relation which HEP holds with all • relevant major Network authorities globally. • 4. To bring the technology of very high rate/long distance • data transport to practical use in experimental production • environments (both current and LHC experiments).

3. Network Sector : Activities 1. Exploitation of the UKLIGHT infrastructure 2. CCLRC SLA effort for Performance monitoring; Diagnostic Services for Grid Operations; and PPNCG Support. 3. Incorporation of high performance transport technology

4. Exploitation of UKLIGHT UKLIGHT Activities Purpose: to ensure that PP benefits from the UKLIGHT infrastructure by collaborating directly in the first “early success project” in conjunction with other disciplines, including Radio Astronomy and possibly the HPC community. to show the benefit of “light-path” switching to LHC and Tevatron and/or BaBar data delivery. It is important to emphasise that GridPP-2 is not expected to resource the entire activity, but, to leverage a larger effort and thereby ensure that the outcome benefits HEP, and that GridPP-2 retains its prominence in this area.

5. Deliverables D2.1 [M6] Static light-path connectivity to CERN and FNAL (dependent upon FNAL fibres) D2.2 [M12] Demonstration of light-path switching services for data transport for at least one experiment (CERN or FNAL) Demonstration of multi Gigabit/s transport with test traffic. D2.3 [M18] Demonstration of multi Gigabit/s (with goal of > 5 Gbit/s) transport with one or more advanced protocols. A significant contribution to at least one showcase venue, including SC2004 and IGRID D2.4 [M24] Demonstration of light-path switching services to connect EGEE/LGC main Tier sites. D2.5 [M30] Demonstration of light-path switching services for at least two experiments, at least one of which must be an LHC experiment D2.6 [M36] Strategic report describing the benefits of hybrid service networking to particle physics.

6. CCLRC SLA Activities #1 Performance monitoring and publication Diagnostic engines for Grid operations Purpose: A framework for measurement and publication of a new pervasive network performance measurement capability at end hosts and throughout the network core will be developed and deployed across EGEE and Geant under EGEE. the architecture is based upon GGF standards and Web/Grid services; for the first time a common scheme will be adopted by applications, Geant and participating national networks. This is being jointly led by the UK and leads to specific GridPP-2 tasks, to contribute to (and in some instances lead) development of GGF standards to take and deploy the scheme at GridPP-2 sites, ensure it is integrated into the UK specific Grid information systems (in addition to EGEE/LCG) and liaise with, and aid, UKERNA to deploy upon SuperJANET; and to develop a diagnostic engine client for the information, for use by Grid operations centres.

7. Deliverables #1 Develop a program to deploy and maintain monitoring tools at PPARC sites and Grid operations centres who wish to receive them. This to include in the later period the diagnostic engine developed under specific deliverables below. Work with the EGEE Network activity JRA4 to ensure all UK performance monitoring deployment (specifically including all relevant GridPP-2 sites) is compliant at equal times with global EGEE deployments according to dates specified by EGEE JRA4 deliverables. D 3.2 [M6] Document detailing diagnostic engine requirements and use- cases as specified primarily by Grid operations centres and EGEE and Internet2 e2e Pipes initiative. D 3.3 [M12] Prototype diagnostic engine satisfying above requirements. This must be deployed and demonstrated within the UK GridPP operations centre. D 3.4 [M18] Document detailing user experience of prototype leading to revised requirements D 3.5 [M24] Refined diagnostic engine satisfying revised requirements. This must be deployed and demonstrated at Grid operations centres.

8. CCLRC SLA Activities #2 Support for PPNCG Purpose: The PPNCG has formal SIG (Special Interest Group) status within the HEFCE/JISC structure (PP, including Astronomy, is the only application area accorded this status). This leads to the PPNCG being consulted at high level and being invited to participate in strategic activities (e.g. SuperJANET 5 planning). To maintain this status requires some modest support to maintain and develop the underlying monitoring tools and GUI presentation, and produce documentation for best practices. Deliverables #2: Provide "expert consultancy" to the GridPP community on all aspects of networking as appropriate (includes an element in support of access to UKLight at RAL) Undertake other tasks as requested by the PPNCG from time to timeincluding liaison activities with providers

9. High Performance Transport High Performance Transport Purpose : to concentrate effort on the very practical aspects of bringing the knowledge which exists in the protocol development domain into practical use by HEP experiments. to demonstration of one or more UK HEP experiments being able to regularly transport data at rates in excess of 500 Mbit/s and preferably 1 Gbit/s. Deliverables D 3.1 [tbd] Provide user support material, including full documentation and "cook book", in respect of users high performance data transport .

10. Reference: EGEE • Services • Deliver “production level” grid services (manageable, robust, resilient to failure) • Middleware • Grid middleware re-engineering activity in support of the production services • Dissemination • Proactively market Grid services to new research communities in academia and industry • Provide necessary education

11. Reference: EGEE Activities Equivalent EDG Work Packages / Groups WP6 WP7 WP1-5 & 6 QAG Security Group WP7 WP12 WP11 WP11 WP8-10 ? • EGEE includes 11 activities • Services • SA1: Grid Operations, Support and Management • SA2: Network Resource Provision • Joint Research • JRA1: Middleware Engineering and Integration • JRA2: Quality Assurance • JRA3: Security • JRA4: Network Services Development • Networking • NA1:Management • NA2:Dissemination and Outreach • NA3: User Training and Education • NA4:Application Identification and Support • NA5:Policy and International Cooperation

12. Grid Middleware Network Resources Reference: EGEE JRA4 • “Network Services Development” • Main sub activities: • Bandwidth Allocation and Reservation • Network Performance Monitoring and Diagnostic tools • IPv6 uptake • GEANT already deploying basic infrastructure: • Introduction of Layer 3 differentiated services • Instrumentation for network measurement • JRA4 Partners • UCL • CNRS / UREC • DANTE • DFN • GARR JRA4 JRA4 will provide interfaces between Grid middleware and Network Resources

13. Reference: EGEE Bandwidth Allocation • EGEE will need to allocate network bandwidth for Grid applications : • immediately and • advance (reservation) • Two features for Bandwidth allocation : • cross multiple domains and, • require Authorization and Authentication for the users.

14. Grid Middleware Grid Users Reference: EGEE Performance Monitoring • Network Performance?, what for? : • Problem diagnostic and rectification • Facilitate resources allocation • Performance monitoring and SLA adherence Grid Ops Network Ops Operations Network Performance Monitoring Network (Dante+NRENs) Common Interface (OGSA) Domain_A Domain_B . . . . Domain_X

15. Ongoing monitoring work • Grid network monitoring services • Strong presence in GGF • NM-WG: Data/control message standardisation • GHPN-RG: Standardised services • Collaboration with Internet 2 (E2E piPES) • Collaboration with Dante (GN2 JRA1) • Measurements across adminstrative domains

16. piPES & Dante work • Both have similar architectures • Measurement points linked to routers on Dante/Geant • Each aims to achieve something quite different • piPES wants a troubleshooting service • Dante want a general informational service

17. GHPN-RG & NM-WG • NM-WG • Specifications for various messages related to network monitoring • GHPN-RG • Describing the services and interactions involved with all sorts of network services • Monitoring, bandwidth allocation, network cost services, network information

18. Long term plans • Full implementation of NMWG schemata • As part of piPES or UK Gridmon • Continue presence within GHPN-RG and any Working Groups spawned from it • Implementations of standards • Extending UK Gridmon or piPES • All very much part of EGEE

19. The Network Sector Questions?