ABSTRACT
Stochastic Rendezvous Networks (SRVNs) are performance models for multitasking parallel software with intertask communication via rendezvous introduced in [1], which are very appropriate to model client-server systems. SRVNs differ from Queueing Networks (QNs) in two ways: nodes act as both clients and servers (allowing for nested service), and servers have two distinct phases of service—the first one “in RV” with the client, and the second “after RV”, executed in parallel with the client. Early work on solving SRVN models has used a kind of approximate Mean Value Analysis based on heuristic ad hoc assumptions to determine the task queue properties at the instant of RV request arrivals. Approximation are necessary since SRVN violates product form. Recently, a more rigorous approach was proposed in [2] for the solution of SRVN models, based on a special aggregation (named “Task-Directed Aggregation” TDA) of the Markov chain model describing the interference of different clients that contend for a single server with FIFO queueing discipline and different service times. The algorithm derived in [2] has the limitation that each client may require only a single class of service. In general, a software server offers a range of services with different workloads and functionalities, and a client may need more than one service. The present paper uses the TDA approach to derive an extended algorithm which allows a client to require any number of services from a server by changing randomly the request class. The new algorithm is incorporated into a decomposition method for models with any number of servers. The SRVN modelling technique is applied to a large case study of a distributed database system, giving insight into the behaviour of the system and helping to identify performance problems such as software bottle-neck.
- 1.C. M. Woodside, "Throughput calculation for basic Stochastic Rendezvous Networks," Performance Evaluation, vol. 9, pp. 143-160, 1989. Google Scholar
Digital Library
- 2.D. C. Petriu and C. M. Woodside, "A new mean value analysis of client-server software by taskdirected aggregation of markov models," Tech. Rep. SCE-93-29, Carleton University, 1993.Google Scholar
- 3.D. Dewire, Client/Server Computing. McGraw- Hill, 1993. Google Scholar
Digital Library
- 4.C. M. Woodside, a. E. Neilson, D. C. Petriu, and S. M ajumdar, "The rendezvous network model for performance synchronous multi-tasking distributed software," Tech. Rep. SCE-89-8, Carleton, 1989.Google Scholar
- 5.D. C. Petriu and C. M. Woodside, "Approximate MVA from M arkov model of software client/server systems," in Proc. of The Third IEEE Syrup. on Parallel and Distr~b. Proc., (Dallas), pp. 322-329, 1991.Google Scholar
- 6.M. Ajmone Marsan, G. Balbo, and G. Conte, "A class of generalized stochastic petri nets for the performance evaluation of multiprocessor systems," A CM TOCS, vol. 2, no. 2, pp. 93-122, 1984. Google Scholar
Digital Library
- 7.J. A. Rolia, "Predicting the performance of software systems," Report CSRI-TR-260, University of Toronto, Canada, 1992.Google Scholar
- 8.P. Heidelberger and K. S. Trivedi, "Analytic queueing models for programs with internal concurrency," IEEE TraT~s. on Computers, vol. C-32, no. 1, pp. 73-82, 1983.Google Scholar
- 9.A. Thomasian and P. Bay, "Analytic queuing network models for parallel processing of task systems," IEEE Trans. on Computers, vol. C-35, no. 12, pp. 1045-1054, 1986. Google Scholar
Digital Library
- 10.A. Kapelnikov, R. R. Muntz, and M. D. Ercegovac, "A methodology for performance analysis of parallel computations with looping constructs," Journal of ParaIlel and Distmb. Computing, no. 14, pp. 105- 120, 1992. Google Scholar
Digital Library
- 11.D. C. Petriu, "Approximate mean value analysis of client-server systems with multi-class requests," Tech. Rep. SCF_,-93-30, Carleton University, October 1993.Google Scholar
- 12.G. Chiola, GreatSPN Users' Manual, Version 1.3. 1987.Google Scholar
- 13.J. E. Neilson, "Parasol: A simulator for distributed and/or parallel systems," Tech. Rep. SCS-TR-192, Carleton University, 1991.Google Scholar
- 14.A. Borr and F. Putzolu, "High performance sql through low-level system integration," in Proc. Int. Conf. on Management of Data (SIG3IOD), pp. 342-349, 1988. Google Scholar
Digital Library
- 15.C. U. Smith, Performance Engineering of Software Systems. Addison-Wesley, 1990. Google Scholar
Digital Library
- 16.J. E. Neilson, C. M. Woodside, D. C. Petriu, and S. M ajumdar, "Software bottlenecking in clientserver systems and rendezvous networks," Tech. Rep. SCE-92-17, Carleton University, 1992.Google Scholar
Index Terms
Approximate mean value analysis of client-server systems with multi-class requests
Recommendations
Approximate mean value analysis of client-server systems with multi-class requests
Stochastic Rendezvous Networks (SRVNs) are performance models for multitasking parallel software with intertask communication via rendezvous introduced in [1], which are very appropriate to model client-server systems. SRVNs differ from Queueing ...
Approximate MVA for Client-Server Systems with Nonpreemptive Priority
MASCOTS '95: Proceedings of the 3rd International Workshop on Modeling, Analysis, and Simulation of Computer and Telecommunication SystemsA new approximate algorithm for the Stochastic Rendezvous Network (SRVN) model with nonpreemptive priority scheduling is introduced in this paper. SRVN is a performance model for client-server systems with synchronous communication which is different ...
Approximate MVA from Markov model of software client/server systems
SPDP '91: Proceedings of the 1991 Third IEEE Symposium on Parallel and Distributed ProcessingStochastic rendezvous networks (SRVN) are performance models for multitasking parallel software with intertask communication via rendezvous (RV). The paper describes an approximation for the arrival instant probabilities for a class of simple client/...






Comments