M. Calzarossa, L. Massari, A. Merlo, M. Pantano, and D. Tessera, "Medea: A Tool for Workload Characterization of Parallel Systems," IEEE Parallel and Distributed Technology, 3(4), pp. 72-80, November 1995.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....Data management and measurement analysis supports a rich set of performance data reduction, filtering, summary, and analysis techniques. Many performance metrics and statistics can be computed. SCALA supports several trace formats which allows the use of various visualization systems (e.g. Medea [7], and Upshot [23] Finally, SCALA has highly sophisticated scalability analysis integrated, that examines the scaling behavior of a program for varying input data and machine sizes. This paper describes the performance oriented development of real world applications for distributed and parallel ....

....programs. SCALA supports several different trace formats including ALOG and Grace formats for collected performance data which enables the usage of various performance visualization systems. Based on the ALOG trace format we can use well established visualization systems such as Medea [7], TAU [36] and Upshot [23] Medea is a post mortem performance analysis and visualization system. Among others we use Medea to derive and visualize performance metrics together with the input program. This work has been described in detail in [8] Figure 5 shows Medea visualizing various ....

M. Calzarossa, L. Massari, A. Merlo, M. Pantano, and D. Tessera. Medea: A tool for workload characterization of parallel systems. IEEE parallel and distributed technology: systems and applications, 3(4):72--80, Winter 1995.

....and ATExpert. In addition, numerous researchers have applied statistical techniques to performance data in an effort to reduce data volume or to automate tasks for the user. These techniques include covariance analysis, discriminant analysis, principle component analysis, and clustering analysis [3, 19]. In the knowledge discovery field, Lee, Stolfo, and Mok [12] have focused techniques for machine learning on traces of Internet network activity to provide automated support for intrusion detection in computer networks. To assist with the analysis of large trace files, numerous researchers have ....

M. Calzarossa, L. Massari et al., "Medea: A Tool for Workload Characterization of Parallel Systems," IEEE Parallel &Distributed Technology, 3(4):72-80, 1995.

....requires capture from remote sites, network transmission, near real time data convolution and remote instrument control. Tuning the behavior of such distributed applications requires deep understanding of network, parallel system and application behavior. Although significant research projects [2, 6] have already applied visualization to understanding software behavior, this work has focused on either single processor or homogeneous parallel systems. The added complexity of a networked environment dictates a new breed of visualization tools. These new tools must support effective ....

Calzarossa, M. et al. "Medea: A Tool for Workload Characterization of Parallel Systems,"IEEE Parallel and Distributed Technology, pp. 72-80, winter 1995.

....Data management and measurement analysis supports a rich set of performance data reduction, filtering, summary, and analysis techniques. Many performance metrics and statistics can be computed. SCALA supports several trace formats which allows the use of various visualization systems (e.g. Medea [12], and Upshot [45] Finally, SCALA has highly sophisticated scalability analysis integrated, that examines the scaling behavior of a program for varying input data and machine sizes. SCALA has been used to analyze and to guide the application development by selectively instrumenting and measuring ....

....which enables the usage of several performance visualization systems. The ALOG format (part of the MPICH implementation) describes time stamped events and is the de facto trace format for MPI. Based on the ALOG trace format we can use visualization systems of well established systems such as Medea [12], TAU [56] and Upshot [45] Medea is a post mortem performance analysis and visualization system. This system has been extended for performance analysis and visualization of HPF programs that are parallelized by VFC compiler. SCALA is used to instrument programs and to generate ALOG trace files. ....

Maria Calzarossa, Luisa Massari, Alessandro Merlo, Mario Pantano, and Daniele Tessera. Medea: A tool for workload characterization of parallel systems. IEEE parallel and distributed technology: systems and applications, 3(4):72--80, Winter 1995.

....Data management and measurement analysis supports a rich set of performance data reduction, filtering, summary, and analysis techniques. Many performance metrics and statistics can be computed. Scala supports several trace formats which allows the use of various visualization systems (e.g. Medea [6], and Upshot [19] Finally, Scala has highly sophisticated scalability analysis integrated, that examines the scaling behavior of a program for varying input data and machine sizes. This paper describes the performance oriented development of real world applications for distributed and parallel ....

....parallel programs. Scala supports two trace formats ALOG and Scala specific format for collected performance data which enables the usage of several performance visualization systems. Based on the ALOG trace format we can use visualization systems of well established systems such as Medea [6], TAU [31] and Upshot [19] Medea is a post mortem performance analysis and visualization system. Among others we use Medea to derive and visualize performance metrics together with the input program based on ALOG trace files. This work has been described in detail in [7] Scala also generates ....

M. Calzarossa, L. Massari, A. Merlo, M. Pantano, and D. Tessera. Medea: A tool for workload characterization of parallel systems. IEEE parallel and distributed technology: systems and applications, 3(4):72--80, Winter 1995.

....increases. However, the large increase in the total number of accesses in Cholesky is due to the increase in both private and shared accesses. 6 Related Work Available parallel performance analysis tools have mainly been developed for analyzing messagepassing applications, e.g. Pablo [18] Medea [19], and Paradyn [20] Nevertheless, there is some work that focuses on characterizing shared memory applications. Singh et al. demonstrated that it is often difficult to model the communication of parallel algorithms analytically [17] They suggested developing general purpose simulation tools to ....

M. Calzarossa, L. Massari, A. Merlo, M. Pantano, and T. Daniele, "Medea: A tool for workload characterization of parallel systems," IEEE Parallel and Distributed Technology, vol. 3, pp. 72--80, Winter 1995.

....characterizing, we are collecting traces using code instrumentation and other methods. Trace collection using code instrumentation is described in this report. Most of the available parallel performance analysis tools are developed for analyzing messagepassing applications, e.g. Pablo [9] Medea [10], and Paradyn [11] The tools presented in this report address the shortage of tools for analyzing shared memory applications. They assist in the performance analysis of shared memory applications and enable characterizing wide range of performance aspects. We have developed two tools for ....

M. Calzarossa, L. Massari, A. Merlo, M. Pantano, and T. Daniele, "Medea: A tool for workload characterization of parallel systems," IEEE Parallel and Distributed Technology, vol. 3, pp. 72--80, Winter 1995.

.... profiling, and tracing support [6] The source level evaluation of C, C , Fortran and HPF codes is addressed by the MPP Apprentice tool [8] and by the Visualization Tool [7] In this paper we present the integration of the VFC compilation system [3] and the Medea performance analysis tool [4]. This work is part of the Esprit LTR project HPF 1 , whose goal is to extend the current version of the HPF language and the related compiling technology to address the requirements of advanced application problems. HPF language includes all HPF 2 features, a few of HPF 2 approved ....

M. Calzarossa, L. Massari, A. Merlo, M. Pantano, and D. Tessera. Medea: A Tool for Workload Characterization of Parallel Systems. IEEE Parallel and Distributed Technology, 3(4):72--80, 1995.

....number of processors or sliding windows of the most recent data. However, to manage large amount of data and large number of per 8 formance indices, advanced statistical techniques need to be applied. Clustering is one of the most common techniques used to reduce the amount of performance data [3, 12] and consequently reduce the number of performance indices needed to identify sources of performance losses. Statistical clustering is a multidimensional technique useful to automatically identify equivalence classes. For example, clustering classifies processors such that all processors in one ....

....System (SIS) SIS [4] is a tool that allows the automatic instrumentation, via command line options, of various code regions such as subroutines, loops, independent loops and any arbitrary sequence of executable statements. SIS has been extensively tested in connection with the Medea tool [3] for investigating the performance of applications with irregular data accesses. SCALA is being integrated within new compilation framework. The data management module accepts tracefiles obtained executing the code instrumented with SIS and computes a set of statistical metrics for each code ....

Calzarossa, M., Massari, L., Merlo, A., Pantano, M., and Tessera, D. Medea: A tool for workload characterization of parallel systems. IEEE Parallel & Distributed Technology Winter (1995), 72--80.

No context found.

M. Calzarossa, L. Massari, A. Merlo, M. Pantano, and D. Tessera, "Medea: A Tool for Workload Characterization of Parallel Systems," IEEE Parallel and Distributed Technology, 3(4), pp. 72-80, November 1995.

No context found.

Maria Calzarossa, Luisa Massari, Alessandro Merlo, Mario Pantano, and Daniele Tessera, Medea:A Tool for Workload Characterization of Parallel Systems,IEEE Parallel & Distributed Technology: Systems & Applications, Vol. 3, No. 4, Winter 1995, pp. 72-80.

No context found.

M. Calzarossa et al., "Medea: A Tool for Workload Characterization of Parallel Systems," IEEE Parallel &Distributed Technology, Winter 1995, pp. 72-80.

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