Conference Paper

Applying Advance Reservation to Increase Predictability of Workflow Execution on the Grid

University of Innsbruck, Austria
DOI: 10.1109/E-SCIENCE.2006.261166 Conference: e-Science and Grid Computing, 2006. e-Science '06. Second IEEE International Conference on
Source: IEEE Xplore

ABSTRACT In this paper we present an extension to devise and implement advance reservation as part of the scheduling and resource management services of the ASKALON Grid application development and runtime environment. The scheduling service has been enhanced to offer a list of resources that can execute a specific task and to negotiatewith the resource manager about resources capable of processing tasks in the shortest possible time. We introduce progressive reservation approach which tries to allocate resources based on a fair-share principle. Experiments are shown that demonstrate the effectiveness of our approach, and that reflect different QoS parameters including performance, predictability, resource usage and resource fairness.

0 Followers
 · 
110 Views
  • Source
    • "We show here how to use the profile to build a conservative backfilling scheduler [7], but it should be straightforward to implement other policies. The example also demonstrates how to obtain the scheduling options for a request so that a scheduler can select resources using approaches such as best-fit and worst-fit to minimise a queue's fragmentation and improve resource utilisation [2]–[4], [17], [18]. "
    [Show abstract] [Hide abstract]
    ABSTRACT: This paper details a data structure for managing and scheduling requests for computing resources of clusters and virtualised infrastructure such as private clouds. The data structure uses a red-black tree whose nodes represent the start times and/or completion times of requests. The tree is enhanced by a double-linked list that facilitates the iteration of nodes once the start time of a request is determined by using the tree. We describe the data structure main features, provide an example of use, and discuss experiments that demonstrate that the average complexity of two operations are often below 10% of their respective theoretical worst cases.
  • Source
    • "As stated before, in the Grid domain advance reservation was early identified as necessary for coallocations and guaranteed finish times [8]. Hence, there are a number of articles covering the negotiation and handling of advance reservation [9], [10], [11], [12], [2], [13], [14]. An example for an advance reservation enabled Grid broker is the Virtual Resource Manager (VRM) [15]. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Advance reservation was identified as a key technology to enable guaranteed Quality of Service and co-allocation in the Grid. Nonetheless, most Grid and local resource management systems still use the queuing approach because of the additional complexity introduced by advance reservation. A planning based resource management system has to keep track of the reservations in the future and needs a good overview on the available capacity during the negotiation of incoming reservations. For advance reservation, the resource management problem becomes a two dimensional problem. In this paper different data structures are investigated and discussed in order to fit to planning based resource management. As a result the benefits of using lists of resource allocation or free blocks are exposed. This general idea widely used to manage continuous resources is extended to cover not only the resource dimension but also the time dimension. The list of blocks approach is evaluated in a Grid level and a resource level resource management system. The extensive simulations showed a better runtime and higher reservation success rate compared with the currently favored approach of a slotted time.
    23rd International Symposium on Computer Architecture and High Performance Computing (SBAC-PAD); 01/2011
  • Source
    • "2.1.2. Advance reservations and creation of alternatives to rejected requests Mechanisms for elastic advance reservations and generation of alternative time slots for advance reservation requests have been proposed [22] [23]. These models can be incorporated in the provisioning scenario described in this work to improve resource utilisation and generate alternative offers for provisioning violations. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Several Grids have been established and used for varying science applications during the last years. Most of these Grids, however, work in isolation and with different utilisation levels. Previous work has introduced an architecture and a mechanism to enable resource sharing amongst Grids. It has demonstrated that there can be benefits for a Grid to offload requests or provide spare resources to another Grid. In this work, we address the problem of resource provisioning to Grid applications in multiple-Grid environments. The provisioning is carried out based on availability information obtained from queueing-based resource management systems deployed at the provider sites which are the participants of the Grids. We evaluate the performance of different allocation policies. In con- trast to existing work on load sharing across Grids, the policies described here take into account the local load of resource providers, imprecise availability information and the compensation of providers for the resources offered to the Grid. In addition, we evaluate these policies along with a mechanism that allows resource sharing amongst Grids. Experimental results obtained through simulation show that the mechanism and policies are effective in redirecting requests thus improving the applications' average weighted response time.
    Information and Software Technology 01/2009; 51(1):42-55. DOI:10.1016/j.infsof.2008.09.013

Preview

Download
3 Downloads
Available from