Group Discussion and for Anatomy of the Grid PaperFor the connectivity layer, which the authors describe "communicating easily andsecurely", most of the attention is given to authentication solutions. Traditional Gridinfrastructure, such as the GSI from Globus, has also concentrated on authentication andnot provided a sufficient infrastructure for the rest of the trust hierarchy (i.e.authorisation, policy implementation, etc). 1. To date, what measures have been made towards providing a more complete security infrastructure on the Grid? Since the paper there are have been works to develop authorization framework in GlobusToolkit Siebenlist et. al. at Argonne. There are also pluggable modules for processing authentication, gathering and processing attributes and rendering decisions. In addition, there has been work in OGSA-Authz WG to allow for callouts to third-party authorization services. E.g. PERMIS, CAS.2. Are any of these solutions also "easy" to install and configure?See Information on GAMAhttp://www-128.ibm.com/developerworks/grid/library/gr-unified3/With respect to storage resources, the authors mention that "third party and highperformance transfers are useful". Grid applications today generate very large datasets,and I/O is already the slowest computational component by several orders of magnitudewhen compared to memory or processor speed.3. As number and size of Virtual Organizations grow, will not data access andretrieval become a major bottleneck, and hence remove the attractiveness of thegrid for meeting the needs of some members? What approaches are being used toaddress problem now rather than waiting for network speeds to improve?See answer on presentation slides, basic consensus is using concepts from distributedcomputing such as parallel I/O, keeping redundant copies of the data.4. What work is being done to address some of the more difficult I/O topics for Gridcomputing such as noncontiguous I/O, caching, and fault tolerance?There have been lots of research in GRID I/O and many approaches have been proposedto solve the addressed issues. Grid DataFarm - provides global parallel file systemdesigned for grid clusters that span over 10,000 nodes. MOCHA - self extensibledatabase middleware for interconnecting distributed data sources. VPFS - VersioningParallel File System, and Direct Access Cache, Armada Parallel File System. Discussioninvolved ensuring that problems from other disciplines are rethought and applied to theGRID to determine if they are still applicable and correct with respect to the Gridenvironment.The authors define the Grid problem as flexible, secure, coordinated resource sharingamong dynamic collections of individuals, institutions, and resources. Thus, Gridcomputing by definition cuts across organizational boundaries. Its development issupported by a multitude of working groups and forums where competitors often findthemselves working together for the common goal and collaboration is the operative tool.The rest of the group discussion was about interoperability and the overall vision of the Grid with respect to Grid standards. In terms of the future of the Grid, one of these standards may eventually become dominant, and once organizations see the need for Gridservices and the benefits many boundaries limiting the unification of the Grid will be removed. Also, see answers to other questions in the presentation