Ruben S. Montero, E. Huedo and I. M. Llorente
Distributed Systems Architecture Group
Facultad de Informatica
Universidad Complutense de Madrid
Email: rubensm@dacya.ucm.es
GridWay(www.gridway.org) is an open-source meta-scheduling technology that, on top of Globus toolkit services, performs job execution management and resource brokering, allowing unattended, reliable, and efficient execution of jobs, array jobs, or workflows on heterogeneous and dynamic Grids. GridWay performs all the job scheduling and submission steps transparently to the end user and adapts job execution to changing Grid conditions by providing fault recovery mechanisms, dynamic scheduling, migration on-request and opportunistic migration. The meta-scheduling technology gives end users and application developers a scheduling framework similar to that found on local DRM systems (e.g. PBS,SGE...) for submitting, monitoring, synchronizing and controlling jobs by means of a DRM-like CLI and the DRMAA API GGF standard.
The presentation describes the features and benefits provided by GridWay, its internal architecture and examples of scheduling infrastructures for enterprise, partner and utility grids. These infrastructures follow the end-to-end principle, which has fostered the spectacular development of the Internet. Clients have access to a wide range of resources provided through a limited, standardized set of protocols and interfaces (TCP/IP). In the Grid, the Globus core grid middleware provides this set.
A. Enterprise Grid Infrastructures
Enterprise grids enable diverse resource sharing to improve internal collaboration and achieve a better return from IT investment. Available resources within a company are better exploited and the administrative overhead is minimized by using Grid technology. Enterprise grid infrastructures require organization-level meta-schedulers to provide support for multiple users in a single scheduling instance, and so a centralized approach for scheduling and accounting.
B. Partner Grid Infrastructures
Partner grid infrastructures aim to provide large-scale, secure and reliable sharing of resources among partner organizations. The multiple administration domains existing in a partner grid infrastructure prevent the deployment of centralized meta-schedulers, with total control over client requests and resource status. Therefore, organization-level meta-schedulers should provide support for multiple intra-organization users in each scheduling instance.
C. Utility Grid Infrastructures
Utility grids should provide access to the latest computing platform and technology and still be flexible enough to adjust capacity as required without needing to purchase costly hardware. The deployment of a utility computing solution involves a full separation between the provider and the consumer. So it should be based on standards and allow a gradual deployment in order to obtain a favourable response from the application developers and the IT staff. Grid technology overcomes utility computing challenges by means of its standard functionality for flexible integration of diverse distributed resources. The technological feasibility of the utility model for computing services can be established by using a grid infrastructure based on Globus Toolkit components and the GridWay meta-scheduler. The main feature of this model is the use of the WS-GRAM service to recursively interface to the services available in a Globus based grid. A WS-GRAM service hosting a GridWay meta-scheduler provides the standard functionality required to implement a gateway to a federated grid. Such a combination allows the required virtualization technology to be created in order to provide a powerful abstraction of the underlying grid resource management services.
