Molecular Structure Determination on a Computational and Data Grid - Parallel Computing for Bioinformatics and Computational Biology - page 602

Figure 24.4 ACDC-Grid Monitoring and Information Services-Core Infrastructure architec-
ture description.
and hierarchical in design, making it extremely scalable with tuneable information
time scales.
Our monitoring system is being enhanced to
1. Provide a programmatic interface to the ACDC Job Monitoring database for
running, queued, or historical jobs, complete with the current site status metrics;
2. Provide integration with MonaLisa [95] and the Grid3 site status catalog for
defining difference metrics on job monitoring, resource utilization, and policy
metrics;
3. Provide integration with the Globus MDS provider development team for
providing XML formatted job information and site status metrics;
4. Provide integration of our predictive scheduling estimates based on resource
policy specifications;
5. Provide resource specific CPU availability for Grid3 resources, ACDC-Grid
resources, and VOs;
6. Provide currently available free nodes and predictive scheduling capabilities
of job execution start times based on running, queued, and submitted job
characteristics, including site policy constraints;
7. Provide data grid historical and near real-time estimates of bandwidth and
utilization of grid-enabled repositories; and
8. Harden the secure, lightweight, scalable distributed hierarchical imbed-
ded MySQL database ACDC-Grid monitoring daemon infrastructure for
heterogeneous computational grid hardware resources and heterogeneous grid
infrastructure middleware.