Download presentation
Presentation is loading. Please wait.
Published byLindsey Burke Modified over 9 years ago
1
M.Kunze, ACAT2003, Tsukuba ACAT2003 Session 1 Summary Talk: Computing Technology and Environment for Physics Research Marcel Kunze Institute for Scientific Computing (IWR) Forschungszentrum Karlsruhe GmbH
2
M.Kunze, ACAT2003, Tsukuba Topics (33 Talks and Round Table Discussion) u Parallel Computing Technologies and Applications (7 Talks) u Data Fabric and Data Management (5 Talks) u Online Monitoring and Control (3 Talks) u Advanced Analysis Environments (8 Talks) u Innovations in Software Engineering (5 Talks) u Graphic User Interfaces, Common Libraries (5 Talks) It is almost impossible to treat everything in this talk: It will be necessary to focus on specific topics.
3
M.Kunze, ACAT2003, Tsukuba Computing Platforms Evolution At the Oberammergau workshop (1993), I had given a summary talk labeled „Computing in the 90‘s“. We saw the transition from mainframes to workstations and I emphasized the importance of the client-server model. Today, platforms and applications are far more u Advanced u Powerful u Dynamic u Complex This talk is about distributed computing and the importance of Grid Services.
4
M.Kunze, ACAT2003, Tsukuba A little bit of History
5
M.Kunze, ACAT2003, Tsukuba Grid Computing u Grid Computing has emerged as an important new field, distinguished from conventional distributed computing by ist focus on large-scale resource sharing, innovative applications, and, in some cases, high- performance orientation. Foster, Kesselman, Tuecke, The Anatomy of the Grid: Enabling Scalable Virtual Organizations, 2001 For us this means: u Provision of common tools, frameworks, environment, data persistency u Exploiting the resources available to experiments in computing centres, physics institutes and universities around the world u Presenting this as a reliable, coherent environment for the experiments u The goal is to enable the physicist to concentrate on science, unaware of the details and complexity of the environment they are exploiting
6
M.Kunze, ACAT2003, Tsukuba The Golden Rule Keep it simple As simple as possible Not any simpler - Einstein
7
M.Kunze, ACAT2003, Tsukuba Middleware u The tools that provide functions n that are of general application.. n …. not HEP-special or experiment-special n and that we can reasonably expect to come in the long term from public or commercial sources (cf internet protocols, unix, html)
8
M.Kunze, ACAT2003, Tsukuba Grid Services
9
M.Kunze, ACAT2003, Tsukuba Open Grid Services Infrastructure (OGSI) u Distributed applications are made of software components u Grid Services are an extension of Web Services n Discovery n Dynamic service creation n Lifetime management n Notification u OGSI defines a set of standardized interfaces and protocols u Currently available OGSI implementations n Unix: Globus Toolkit 3, OGSI::Lite (Perl), pyGridWare (Python) n Windows: OGSI.NET (Virginia Univ.); MS.NETGrid (EPCC) u GT3 has been evaluated by LCG (Talk: M. Lamanna) n Generally impressed with GT3 and the overall concept n GT3 IndexService: totally new, looks well designed n Information system and GRAM (critical parts of the GLOBUS kit) have problems of scalability and reliability
10
M.Kunze, ACAT2003, Tsukuba Data Grids: Architecture u Replica Consistency Service in a Data Grid (Talk: G.Pucciani ) n Replication of data increases system performance n Problem: Consistency management is an issue in applications where users can modify replicas u Integrating SRB with the GIGGLE/EDG framework (Talk: S.Metson) n Active collaboration between members of CMS, BaBar and the SDSC SRB Group n Problem: Could files stored in SRB be accessed by LCG tools? n Data discovery component is well understood n Full interoperation requires further development effort n Implementation of corresponding Grid services is planned
11
M.Kunze, ACAT2003, Tsukuba Grid Tools: Monitoring u Configuration Monitoring Tool for Large Scale Distributed Computing (Talk: Y.Wu) n track and query site configuration information for large-scale distributed CMS applications n Plans to rework the tool as a Grid service
12
M.Kunze, ACAT2003, Tsukuba Grid Tools: Monitoring u MonALISA: Monitoring Agents using a Large Integrated Services Architecture (Talk: I.Legrand) n Dynamic registration and discovery & subscription mechanism n Adaptability and self-organization
13
M.Kunze, ACAT2003, Tsukuba Distributed Systems: Simulation u MONARC simulation framework (Talk: I.Legrand) n Modelling of large scale distributed computing systems n Design tool for large distributed systems n Performance evaluation
14
M.Kunze, ACAT2003, Tsukuba Distributed Physics Data Analysis u Most HEP experiments are developing frameworks for distributed computing (M.Burgon-Lyon, G.Garzoglio:CDF,D0; P.Elmer,A.Hasan:BaBar; I.Adachi, G.Moloney:Belle; L.Taylor:CMS; A.J.Peters:ALICE) n Various workable solutions exist n Sometimes parallel and non-compatible effort n Importance of standardization u ARDA: Architectural Roadmap towards Distributed Analysis RTAG11: http://www.uscms.org/s&c/lcg/ARDA/ RTAG11 n Common Grid analysis architecture for all LHC experiments n OGSI compliant n Concern: Analysis activities require chaotic access to resources by a large number of potentially inexperienced users (Professors) n Component-by-component deployment and avoiding big-bang releases are critical parts of the implementation strategy n Recommendation: Prototype based on AliEn (Talk: A.J.Peters)
15
M.Kunze, ACAT2003, Tsukuba Advanced Analysis Environments
16
M.Kunze, ACAT2003, Tsukuba Advanced Analysis Environments
17
M.Kunze, ACAT2003, Tsukuba Advanced Analysis Environments
18
M.Kunze, ACAT2003, Tsukuba General Re-Use of Components and Services (95%)
19
M.Kunze, ACAT2003, Tsukuba Interactive Physics Data Analysis u Issues n Typical interactive requests will run on o(TB) distributed data n Transfer/replication times for the whole data about one hour n Data transfers once and in advance of the interactive session n Allocation, installation and set-up of corresponding database servers before the interactive session n Integration of user-friendly interactive access
20
M.Kunze, ACAT2003, Tsukuba Interactive Physics Data Analysis
21
M.Kunze, ACAT2003, Tsukuba Parallel ROOT Facility: PROOF Local Remote Selection Parameters Procedure Proc.C PROOF CPU TagD B RD B DB 1 DB 4 DB 5 DB 6 DB 3 DB 2 Talk: Fons Rademakers
22
M.Kunze, ACAT2003, Tsukuba PROOF: Actual Development
23
M.Kunze, ACAT2003, Tsukuba PEAC System Overview
24
M.Kunze, ACAT2003, Tsukuba Common Libraries: SEAL u SEAL Project Overview (Talk: L.Moneta) n SEAL has delivered basic foundation, utility libraries and object dictionary n The first version of the Component Model and Framework services is available n Scripting based on Python
25
M.Kunze, ACAT2003, Tsukuba Common Libraries: PI u Physics Interface Project, Status and Plan (Talk: A.Pfeiffer) n Analysis Services components written in Python n Prototypes available to implement AIDA interface for HippoDraw and ROOT
26
M.Kunze, ACAT2003, Tsukuba Graphic User Interfaces: QtRoot u Cross-platform approach to create the interactive application based on ROOT and Qt GUI libraries (Talk: V.Fine) n Qt package from TrollTech AS is a multi-platform C++ application framework that developers can use to write single-source applications that run-natively-on Windows, Linux, Unix, Mac OS X and embedded Linux. n A lot of Qt widgets available for re-use n Qt is the basis for the KDE desktop n Consolidation of Root Graphics (TGQt vs. TGWin32,TGX11,TGWin32GDK) Example: A fragment of STAR “Event Display” QtGLViewer class based viewer see: http://www.rhic.bnl.gov/~fine/EventDisplay )http://www.rhic.bnl.gov/~fine/EventDisplay
27
M.Kunze, ACAT2003, Tsukuba Fabric Area
28
M.Kunze, ACAT2003, Tsukuba Data Fabric and Data Management u Need for powerful, high throughput systems u Storage Area Networks n GridKa scalable IO design based on fibre channel technique (Talk: J.van Wezel) n Infiniband yields 800 MB/s (Talk: U. Schwickerath) u Need for powerful trigger systems to reduce data n Realtime analysis for the ALICE HLT (Talk: C.Loizides) u Need for powerful clusters and networks for online event reconstruction and distributed analysis n Realtime event reconstruction farm for Belle (Talk: R.Itoh) n A basic R&D for an analysis framework distributed on wide area network (Talk: H. Sakamoto) u New methods for data integration and management n Grid portal based data management for lattice QCD data (GENIUS Talk: G.Andronico)
29
M.Kunze, ACAT2003, Tsukuba Round Table Discussion D. Laforenza
30
M.Kunze, ACAT2003, Tsukuba
32
It took 200 Years to develop electrical Grids
33
M.Kunze, ACAT2003, Tsukuba Open Questions u Is the far-reaching vision offered by Grid Computing obscured by the lack of interoperability standards among grid computing technologies ? u Should the next few years be considered as a transition period with multiple prototypes in competition to speed up the development ?
34
M.Kunze, ACAT2003, Tsukuba How to design Grid-aware Applications? u Make developers and users aware of network based applications u Need to think about new abstract programming models u Development of new programming techniques and tools that specifically address the Grid and encompass n Heterogeneity n Distributed computing aspects of Grid programming
35
M.Kunze, ACAT2003, Tsukuba CrossGrid: Tools for easy Use of the Grid
36
M.Kunze, ACAT2003, Tsukuba CrossGrid: Migrating Desktop u Idea n Save and resume a user grid session n Look and feel of a windows desktop u Implementation n Roaming Access Server and Clients n Java Web Services (Portability) u Integration of Tools n Job submission wizard n Job monitoring dialog n GridExplorer dialog n GridCommander dialog
37
M.Kunze, ACAT2003, Tsukuba Outlook u Scaling of Fabric Infrastructure n Cheap commodity components vs. High-tech solutions (e.g. SAN) n Note: Each service needs an operator n Total cost of ownership has to take into account infrastructure & manpower n What will be the business model for the Grid market place of resources ? Unlimited access ? Credit points ? Cash ? u ARDA prototype will push development of Physics applications in a distributed environment n What will the production environment look like? n Components will be based on Grid Services n Open Grid Service Infrastructure is the common denominator n Rapid prototyping and user feedback is essential ! n Concern: Users only change their paradigm of working if they see added value (better results, faster turn-around, additional resources etc.) !
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.