“…One utilises underlying continuous time Markov chain (CTMC) properties of queueing models. Additionally, classes of product-form models exist, where state equilibrium probability is a scaled product of the marginal state probabilities of Markov processes that represent individual system components [30]. Therefore, queueing models approximate modern communication systems and their long-term behaviour, without the state explosion problem limiting modelling possibilities.…”
Abstract. Response times are arguably the most representative and important metric for measuring the performance of modern computer systems. Further, service level agreements (SLAs), ranging from data centres to smartphone users, demand quick and, equally important, predictable response times. Hence, it is necessary to calculate moments, at least, and ideally response time distributions, which is not straightforward. A new moment-generating algorithm for calculating response times analytically is obtained, based on M/M/1 processor sharing (PS) queueing models. This algorithm is compared against existing work on response times in M/M/1-PS queues and extended to M/M/1 discriminatory PS queues. Two real-world case studies are evaluated.
“…One utilises underlying continuous time Markov chain (CTMC) properties of queueing models. Additionally, classes of product-form models exist, where state equilibrium probability is a scaled product of the marginal state probabilities of Markov processes that represent individual system components [30]. Therefore, queueing models approximate modern communication systems and their long-term behaviour, without the state explosion problem limiting modelling possibilities.…”
Abstract. Response times are arguably the most representative and important metric for measuring the performance of modern computer systems. Further, service level agreements (SLAs), ranging from data centres to smartphone users, demand quick and, equally important, predictable response times. Hence, it is necessary to calculate moments, at least, and ideally response time distributions, which is not straightforward. A new moment-generating algorithm for calculating response times analytically is obtained, based on M/M/1 processor sharing (PS) queueing models. This algorithm is compared against existing work on response times in M/M/1-PS queues and extended to M/M/1 discriminatory PS queues. Two real-world case studies are evaluated.
“…In addition, we compared our solution method with current RCAT implementations, mainly AutoCAT (Casale and Harrison, 2013) and INAP (Balsamo et al, 2010a;Marin and Bulo, 2009). AutoCAT (Casale and Harrison, 2013) formulates the RCAT sufficient conditions for product-form into a non-linear optimization problem with nonconvex quadratic constraints.…”
Section: Scalabilitymentioning
confidence: 99%
“…In addition, we compared our solution method with current RCAT implementations, mainly AutoCAT (Casale and Harrison, 2013) and INAP (Balsamo et al, 2010a;Marin and Bulo, 2009). AutoCAT (Casale and Harrison, 2013) formulates the RCAT sufficient conditions for product-form into a non-linear optimization problem with nonconvex quadratic constraints. INAP (Balsamo et al, 2010a;Marin and Bulo, 2009) is a fixed point iterative technique to compute the stationary distributions of product-form models that satisfy MARCAT (Harrison and Lee, 2005).…”
Section: Scalabilitymentioning
confidence: 99%
“…Both methods are limited to models with pairwise synchronizations and thus are only suitable for the RB-n-m models when m = 2. Comparing against an open model comprising a BB-2 and three queues (equivalent to 13 equations), our method provided a more efficient solution time of 0.09 s, with AutoCAT and INAP taking 35 and 234 seconds respectively (Casale and Harrison, 2013).…”
Section: Scalabilitymentioning
confidence: 99%
“…Even though these algorithms have not been automated, the implementation is straightforward and can be generalized to identifying reversed rates and the RCAT rate equations from XML representations of Markovian models. • Automated algorithms to compute the values of the reversed rates by applying RCAT conditions to the CTMCs of the cooperating agents (Casale and Harrison, 2013;Balsamo et al, 2010a;Marin and Bulo, 2009). This is conducted without direct use or solution of the rate equations.…”
Section: Appendix Product-forms Using Rcatmentioning
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.