A standards‐based Grid resource brokering service supporting advance reservations, coallocation, and cross‐Grid interoperability

Abstract

SUMMARYThe problem of Grid-middleware interoperability is addressed by the design and analysis of a featurerich, standards-based framework for all-to-all cross-middleware job submission. The service implements a decentralized brokering policy, striving to optimize the performance for an individual user by minimizing the response time for each job submitted. The architecture is designed with focus on generality and flexibility and builds on extensive use, internally and externally, of (proposed) Web and Grid services standards such as WSRF, JSDL, GLUE and WS-Agreement. The external use provides the foundation for easy integration into specific middlewares, which is performed by the design of a small set of plugins for each middleware. Currently, plugins are provided for integration into Globus Toolkit 4, NorduGrid/ARC, and LCG2. The internal use of standard formats facilitates customization of the job submission service by replacement of custom components for performing specific well-defined tasks. Most importantly, this enables the easy replacement of resource selection algorithms by algorithms that addresses the specific needs for a particular Grid environment and job submission scenario. The algorithms in our implementation perform resource selection based on performance predictions, and provide support for advance reservations as well as coallocation of multiple resources for simultaneous job start. The performance of the system is analyzed with focus on overall job service throughput (up to over 250 jobs/minute) and individual job response time (down to under one second).

From volunteer to trustable computing: Providing QoS-aware scheduling mechanisms for multi-grid computing environments

“…As a consequence, self-tuning mechanisms [18] appear as a solution in order to avoid as much as possible those unexpected situations and improve the forecasting, but introducing a general overhead. An algorithm to perform resource selection based on performance predictions, as well as though co-allocation of multiple resources is presented in [56] where already made reservations are displaced based on making co-allocation of jobs. But, unlike our work, those reschedulings are not carried out among jobs belonging to different users.…”

Federating Advanced Cyberinfrastructures with Autonomic Capabilities

“…Among the grid federation efforts we can find InterGrid [9] along with the work by Assuncao et al [3] that promotes interlinking different grid systems through economic-based peering agreements to enable inter-grid resource sharing, Gridway [67] through its grid gateways [22] along the work by Leal et al [35] that proposed a decentralized model for scheduling on federated grids to improve makespan and resource performance, LAGrid meta-scheduling [5,55,59] that promotes interlinking different grid systems through peering agreements to enable inter-Grid resource sharing, Koala [39] with the use of delegated matchmaking [24] to obtain the matched resources from one of the peer Koala instances, VIOLA [61] that implements grid interoperability via WS-Agreement [2] and provides coallocation of multiple resources based on reservations, Grid Meta-Brokering Service (GMBS) [28,29] proposes an architecture for grid interoperability based on high level abstractions to describe the broker's capabilities and properties using a specific language [30][31][32]57], the work by Elmroth et al [13] that presents a grid resource brokering service based on grid standards, Guim et al [56] studied scheduling techniques for multi-site grid environments, and within EGEE, efforts to enable interoperability between gLite and UNICORE [14] systems [38,51].…”

An Advance Reservation-Based Co-allocation Algorithm for Distributed Computers and Network Bandwidth on QoS-Guaranteed Grids

“…Elmroth and Tordsson also propose a co-allocation algorithm [17] for NorduGrid [26]. Their algorithm fixes a reservation time and searches a combination of required computers first, and the related network paths next.…”