This summary is a work-in-progress. Please send comments and suggestions for improvement to childers@mcs.anl.gov
Area | Requirement | Benefits |
---|---|---|
Support for heterogeneous environments | Resource virtualization (i.e., the encapsulation of a diverse implementation behind a common interface) |
|
Common management capabilities (i.e., mechanisms for uniform and consistent resource management) |
Simplifies administration costs | |
Attribute-based resource discovery and query | Enables interactions with resources without knowledge of implementation details | |
Standard protocols and schemas | Facilitates interoperability and wide-spread adoption | |
Area | Requirement | Benefits |
Cross-organizational resource sharing | Global namespace | Enables resource access without regard to location |
Metadata services | Enables accessibility, propagation, aggregation and management of infrastructure across administrative domains | |
Site autonomy | Enables the enforcement of local controls and policies on shared resources | |
Resource usage data | For support of accounting and billing of resource consumption | |
Area | Requirement | Benefits |
Optimization of resource usage | Dynamic resource allocation | Effective supply-side management of resource usage |
Dynamic workload prioritization | Effective demand-side management of resource requests | |
Area | Requirement | Benefits |
Quality of Service Assurance | Service level agreement and attainment | Enables requesters and providers to negotiate terms of use |
Migration | Adjust workloads on demand by migrating executing services and applications | |
Area | Requirement | Benefits |
Administration | Policy-based management mechanisms to ensure operations conform to the goals of the resource owners | To faciliate the automation of system maintainence, to ensure site autonomy |
Problem determination mechanisms | To recognize and efficiently address emerging operational problems | |
Scalable management architecture | To facilitate the deployment, configuration and maintence of complex systems | |
Area | Requirement | Benefits |
High Availability | Disaster recovery mechanisms | Minimizing service disruption in the face of natural or human-triggered disaster |
Mechanisms for fault management, i.e., monitoring, fault detection, diagnostics, etc. | In order that running jobs are not lost due to resource faults | |
Area | Requirement | Benefits |
Job Execution | Support for a variety of remote job types, e.g., simple execution, parallel jobs and more complex workflows | Flexible, on-demand execution of domain-specific services and applications |
Remote job management, e.g., providing the ability to cancel, restart, etc. | Enhanced manageability of domain-specific services and applications | |
Job scheduling | Enables coordination of resources across administrative domains | |
Resource provisioning, including automating the process of resource allocation, configuration and deployment | Enables dynamic and adaptive behavior | |
Area | Requirement | Benefits |
Data Services | Abstractions for data that provide uniform access and integration to various types of data (databases, files, streams) regardless of its physical location. | Presents functionality without requiring knowledge of implementation details |
Mechanisms to keep data consistent across replicas and caches | Ensures data integrity | |
Mechanisms to persist data and associated metadata for its intended lifetime | In support of fault tolerance | |
Mechanisms for data location management, providing ways to transfer, copy and cache data | In support of fault tolerance, performance and scalability | |
Area | Requirement | Benefits |
Security | Authentication and authorization mechanisms | Establish identity and enforce local policies |
Support for multiple security infrastructures | Facilitates integration with existing security infrastructure, supports resource virtualization | |
Perimeter security mechanisms that support local security infrastructure (such as firewall and intrusion-detection policies) while at the same time enabling cross-organizational interactions | Support for site autonomy | |
Isolation of users, performance isolation, isolation between content offerings | Support for site autonomy | |
Delegation of access rights from service requestors to service providers | In support of resource virtualization and site autonomy | |
Support for dynamic negotiation of security policies between service requestors and providers | Support for site autonomy | |
Monitoring in support of intrusion detection, protection and secure logging | Support for site autonomy |
The Anatomy of the Grid
The Physiology of the Grid
The Open Grid Services Architecture, Version 1.0
childers@mcs.anl.gov
Updated 3 July 2006