K. Sreenivasan and A. Kleinman. On the construction of a representative synthetic workload. Communications of the ACM, 17(3):127--133, Mar. 1974.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....the amount of computation and I O were chosen to correspond to actual jobs. Comparison with the use of an actual job stream verified the feasibility of this approach. An elaboration on the use of this idea involves the use of accounting data to provide job resource utilization characteristics (Sreenivasan and Kleinman 1974). In this case, the mix of jobs was chosen through a probability distribution technique with a limit on the total BT time. A second enhancement to this technique has also been implemented. It involves the automatic generation of a complete synthetic job, including job control statements, from ....

Sreenivasan, K., and Kleinman, A. J. (1974). "On the construction of a representative synthetic workload," Communications of the ACM 17(3), 127-33.

....Synthetic programs attempt to combine the advantages of trace data and benchmarks by adding parameters that can be adjusted to reflect real workloads. A discussion of synthetic workload design methods can be found in [22] A few studies using synthetic workloads are found in [23] 24] and [25], which used synthetic workloads to model a file update process. Much of the previous work concerning synthetic workload generation involves workload characterization. In the literature, workload characterization usually refers to the extraction of parameters describing a representative workload, ....

K. Sreenivasan and A. J. Kleinman, "On the construction of a representative synthetic workload," CACM, vol. 17, pp. 127--133, March 1974.

....for centralized systems is a stabilized topic that has been extensively studied and applied since early seventies. Such systems represent the first application domain of workload characterization and many papers on this topic have appeared in the literature since then (e.g. 6] [7], 8] 9] 10] 11] 12] 13] 14] 15] In subsections 3.1 and 3.2, a survey of a few studies, where the various approaches and the techniques described in Section 2 have been applied, will be presented. 3.1 Batch and interactive systems The most popular method adopted for workload ....

K. Sreenivasan and A.J. Kleinman. On the Construction of a Representative Synthetic Workload. Comm. ACM, 17(3):127--133, 1974.

....(say) the particular cache coherence protocol being evaluated. Figure 12 shows the schematic of a workload generator without feedback. In this case, the workload generated does not depend upon the mechanism being measured or modeled. Most existing synthetic workload generators follow this paradigm [48, 63, 74, 98, 139, 168]. While adequate for their specific applications, such generators are not appropriate for our workload generation problem because of the feedback between test jobs and background workload. 40 RESOURCESCHEDULING ALGORITHM MECHANISM MEASURED OR MODELED SYNTHETIC WORKLOAD GENERATOR Figure 12. ....

K. Sreenivasan and A. J. Kleinman, "On the Construction of a Representative Synthetic Workload, " Communications of the ACM, vol. 17, no. 3, pp. 127-132, ACM, 1974.

No context found.

K. Sreenivasan and A. Kleinman. On the construction of a representative synthetic workload. Communications of the ACM, 17(3):127--133, Mar. 1974.

No context found.

K. Sreenivasan and A. J. Kleinman. On the construction of a representative synthetic workload. Communications of the ACM, 17(2):127--133, March 1974.

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