F. Chang, V. Karamcheti, and Z. Kedem. Exploiting application tunability for efficient, predictable parallel resource management. In Proc. 13th Intl. Parallel Processing Symposium, 1999.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....For example, assume ther is an application can be running with 8 way parallelism, unfortunately, during the course of the execution, two processors are shut down. As such, this application must tune itself to be run with 6 processors, result in the delay of total execution time. Chang et.al in [8] study this problem in detail. Whatever, how to describe the tunability of an application does require further research in order to provide a convenient programming model. 4 Related Work Software DSM systems have been widely studied in the past 14 years, M. Rasit Eskicioglu of University of ....

F. Chang, V. Karamcheti, and Z. Kedem. Exploiting application tunability for efficient predictable parallel resource management. In Proc. of the Second Merged Symp. IPPS/SPDP

....explain it here in order to make the paper self contained. A set of t independent tasks is to be mapped onto a system of m machines with the objective of minimizing the total completion time of the tasks. It is assumed that execution time for each task on each machine is known prior to execution [3,8,9,17] and contained within an ETC (Expected Time to Compute) matrix. Each row of the ETC matrix contains the estimated execution times for a given task on each machine. Similarly, each column of the ETC matrix consists of the estimated execution times for each task on a given machine. Thus, ETC[i,j] is ....

F. Chang, V. Karamcheti, Z. Kedem. Exploiting Application Tunability for Efficient, Predictable Parallel Resource Management. Proceedings of the 13

No context found.

F. Chang, V. Karamcheti, and Z. Kedem. Exploiting application tunability for efficient, predictable parallel resource management. In Proc. 13th Intl. Parallel Processing Symposium, 1999.

....are as well as mechanisms for effecting transitions among the various alternatives. We meet these requirements by relying on a structuring concept called the application tunability interface, which was introduced in earlier work done in the context of the Calypso parallel programming language [8]. The tunability interface provides language support to express the availability of multiple execution paths for the application, as well as providing means by which application progress can be monitored and influenced (by switching to a different path) In the context of distributed applications ....

F. Chang, V. Karamcheti, and Z. Kedem. Exploiting application tunability for efficient, predictable parallel resource management. In Proc. 13th Intl. Parallel Processing Symposium, 1999.

....this requires some applicationindependent way of enumerating the different configurations as well as mechanisms for effecting transitions among the various alternatives. We meet these requirements by relying on a structuring concept called the application tunability interface first introduced in [8]. The tunability interface provides language support to express the availability of multiple execution paths for the application, as well as the means by which application progress can be monitored and influenced (by switching to a different path) 3.1. Tunability Interface In our framework, ....

F. Chang, V. Karamcheti, and Z. Kedem. Exploiting application tunability for efficient, predictable parallel resource management. In Proc. 13th Intl. Parallel Processing Symposium, 1999.

....middleware application application application Figure 1. The MILAN middleware in context. dictability, even when the underlying set of machines is unreliable and changing dynamically. The principal outcomes of the MILAN project are (1) a core set of fundamental resource management techniques [21, 3, 23, 9] enabling construction of efficient, reliable, predictable virtual machines, and (2) the realization of these techniques in three complete programming systems: Calypso [3] Chime [27] and Charlotte [6] Calypso extends C with parallel steps interleaved into a sequential program. Each parallel ....

....as image recognition, virtual reality, and media processing that are increasingly running on metacomputers. One of the key challenges deals with providing sufficient resources to computations to enable them to meet their time deadlines in the face of changing resource availability. Our technique [9] relies upon an explicit specification of application tunability, which refers to an application s ability to absorb and relinquish resources during its lifetime, possibly trading off resource requirements versus quality of its output. Tunability provides the freedom of choosing amongst multiple ....

F. Chang, V. Karamcheti, and Z. M. Kedem. Exploiting application tunability for efficient, predictable parallel resource management. In Proceedings of International Parallel Processing Symposium (IPPS/SPDP), 1999.

....execution at run time, respectively. Section 7 describes the use of these components for dynamic adaptation of the visualization application. Related work is discussed in Section 8 and we conclude in Section 9. 2 Application Tunability Application tunability, a notion we initially introduced in [9, 8], refers to an application s ability to trade off resource requirements over several dimensions, including time, quality, and resource type, while still producing an output of adequate quality. Tunable applications are able to compensate for a lower allocation 2 in one stage of the computation by ....

....execution environment, control parameters, application qualities and their measurement mechanism, tunable application modules, and the special points where configuration and reconfiguration can take effect. A detailed description of the syntax and semantics of these annotations appears elsewhere [9, 8]; here, we illustrate their usage using the active visualization application. 4.1 Tunability in Active Visualization Figure 2 shows the specification of alternate configurations in the active visualization application. Although the application has both a server side and a client side component, ....

[Article contains additional citation context not shown here]

F. Chang, V. Karamcheti, and Z. Kedem. Exploiting application tunability for efficient, predictable parallel resource management. In Proc. 13th Intl. Parallel Processing Symposium, 1999.

....execution at run time, respectively. Section 7 describes the use of these components for dynamic adaptation of the visualization application. Related work is discussed in Section 8 and we conclude in Section 9. 2 Application Tunability Application tunability, a notion we initially introduced in [9, 8], refers to an application s ability to trade off resource requirements over several dimensions, including time, quality, and resource type, while still producing an output of adequate quality. Tunable applications are able to compensate for a lower allocation 2 in one stage of the computation by ....

....execution environment, control parameters, application qualities and their measurement mechanism, tunable application modules, and the special points where configuration and reconfiguration can take effect. A detailed description of the syntax and semantics of these annotations appears elsewhere [9, 8]; here, we illustrate their usage using the active visualization application. 4.1 Tunability in Active Visualization Figure 2 shows the specification of alternate configurations in the active visualization application. Although the application has both a server side and a client side component, ....

[Article contains additional citation context not shown here]

F. Chang, V. Karamcheti, and Z. Kedem. Exploiting application tunability for efficient, predictable parallel and distributed systems. Invited submission to Journal of Parallel and Distributed Computing.

.... dynamically changing set of machines applications written for reliable metacomputers virtualization layer Internet MILAN run time middleware application application application The principal outcomes of the MILAN project are (1) a core set of fundamental resource management techniques [21, 3, 23, 9] enabling construction of efficient, reliable, predictable virtual machines, and (2) the realization of these techniques in three complete programming systems: Calypso [3] Chime [27] and Charlotte [6] Calypso extends C with parallel steps interleaved into a sequential program. Each parallel ....

....as image recognition, virtual reality, and media processing that are increasingly running on metacomputers. One of the key challenges deals with providing sufficient resources to computations to enable them to meet their time deadlines in the face of changing resource availability. Our technique [9] relies upon an explicit specification of application tunability, which refers to an application s ability to absorb and relinquish resources during its lifetime, possibly trading off resource requirements versus quality of its output. Tunability provides the freedom of choosing amongst multiple ....

F. Chang, V. Karamcheti, and Z. M. Kedem. Exploiting application tunability for efficient, predictable parallel resource management. In Proceedings of International Parallel Processing Symposium (IPPS/SPDP), 1999.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC