D. A. Menasce, V. A. F. Almeida, R. Fonseca, and M. A. Mendes. Resource management policies for E-commerce servers. In Proc. ACM SIGMETRICS Workshop on Internet Server Performance, 1999.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....been able to acquire the data for the study. 2 Related Work Although there has been several works reported on the workload characetrization of general web servers [4, 8, 14, 20] only a few studies have been reported on the characterization of e commerce trac based on the client behavior. In [9] the authors have developed a resource utilization model for a server which represents the behavior of groups of users based on their usage of the site. This model represents the usage of each resource in the web server. The states of the model can be de ned as a speci c operation performed by a ....

D. A. Menasce, A. A. F. Almeida, R. Fonseca, and M. A. Mendes, \Resource management policies for e-commerce servers," 2nd Workshop on Internet Server Performance, May 1999.

....as the workload generated at a higher level will result in a stream of workload requests at a lower level. Examples for hierar chical user models can be found in [13, 3, 6] Another popular method for catching inter command dependencies is to apply Markov Chains for user behavior mod eling [9, 2, 1, 4]. In [5] users are represented by so called user equivalents , which are simple two level heavy tailed On Off processes. An advanced hierar chical application model has been implemented in the version 7 of the leading industrial network simulator OPNET [11] Although having comparable features, ....

D. Menasce, V. Almeida, R. Fonseca, M. Mendes, Resource management policies for e-commerce servers. In The 2nd Workshop on Internet Server Performance (WISP 99), 1999.

....of representative data. E commerce sites have highly secure information in the traces and access logs. Due to the security implications e commerce sites are reluctant to divulge this information for research purposes. Due to this, studies in this field are still in the preliminary stages. In [8] the authors have developed a resource utilization model for a server which represents the behavior of groups of users based on their usage of the site. It should also be noted that the existing work reported on e commerce traffic has been done on the front end servers only and to the best of our ....

D. A. Menasce, A. A. F. Almeida, R. Fonseca, and M. A. Mendes, "Resource management policies for e-commerce servers," 2nd Workshop on Internet Server Performance, May 1999.

....as the workload generated at a higher level will result in a stream of workload requests at a lower level. Examples for hierarchical user models can be found in [12,2,6,7] Another popular method for catching inter command dependencies is to apply Markov Chains for user behavior modeling [1,3,9]. In [4] users are represented by so called user equivalents , which are simple two level heavy tailed On Off processes. An advanced hierarchical application model has been implemented in the version 7 of the leading industrial network simulator OPNET [11] Although having comparable features, ....

D. Menasce, V. Almeida, R. Fonseca, M. Mendes, Resource management policies for ecommerce servers. In The 2nd Workshop on Internet Server Performance (WISP 99), 1999.

....as the workload generated at a higher level will result in a stream of workload requests at a lower level. Examples for hierarchical user models can be found in [12] and [3] Another popular method for catching inter command dependencies is to apply Markov Chains for user behavior modeling ([9], 2] 1] 4] More static approaches can be found in [11] and [13] where users are described by n dimensional vectors. In [5] users are represented by so called user equivalents , which are simple two level heavy tailed On Off processes. THE BISANTE WORKLOAD GENERATOR The workload of a ....

D. Menasce, V. Almeida, R. Fonseca, M. Mendes, Resource management policies for e-commerce servers. In The 2nd Workshop on Internet Server Performance (WISP 99), 1999.

.... client access patterns, we either use a real trace (in the case of on line forums) or de INFOCOM 2001 7 velop a state transition graph to model client behaviors (in the case of auction and news service) The state transition graph approach was used in a recent study on e commerce user modeling [14]. A. Online Forums We have developed an application to mimic the design and behavior of Web site www.melissavirus.com for online discussion of Melissa virus. There are mainly three types of dynamic pages generated at this Web site: page default.asp is the front page and shows a summary of major ....

....When the content related to a topic (such as sports) changes, all cached pages which contain related headlines (i.e. URL class: cgibin news topic=sports) are invalidated. To model client access patterns at a customized news site, we have followed the state transition graph approach used in [14]. Figure 5 shows our client behavior model. A client moves between two states: SLEEP and STAY. It moves from SLEEP state to STAY state by issuing a new request. During the STAY state, the client may automatically reload the same page every TRE seconds, or with a probability p, the client may ....

Daniel Menasce, Vigilio Almeida, Rodrigo Fonseca, and Marco A. Mendes, "Resource Management Policies for ECommerce Servers," in 2nd Workshop on Internet Server Performance, May 1999.

....the requested content. Similar Web server prototypes that support differentiated services have been proposed in [10, 19] To enforce SLA constraints, Pandey et al. 14] examine selective allocation of server resources through the assignment of different priorities to page requests. Menasce et al. [12] analyze and compare policies that dynamically assign priorities to customers of a commercial Web site by differentiating between visitors and potential buyers. Most of the previous results consider Web sites consisting of a single server node. On the other hand, we claim that popular Web sites ....

D. A. Menasce, J. Almeida, R. Fonseca, and M. A. Mendes. Resource management policies for e-commerce servers. In Proc. Workshop on Internet Server Performance 1999, Atlanta, GE, May 1999.

....customers who pay more money will expect to receive extra quality guarantees. Thus it is essential for an e business site to provide differentiated service quality guarantees that can be geared to make more efficient utilization of available resources. In fact, an analysis by Menasce et al. [6] for an e commerce site shows that as much as 43 revenue increase can be expected if computing resources are allocated based on client states and request semantics. The importance of providing differentiated service qualities has been recognized in the networking community and the focus of these ....

Daniel Menasce, Vigilio Almeida, Rodrigo Fonseca, and Marco A. Mendes, "Resource Management Policies for ECommerce Servers," in 2nd Workshop on Internet Server Performance, May 1999.

....need to be invalidated every 30 seconds. We assume the probability that a topic change contents follows Zipf distribution, a distribution which has been observed in many human behaviors and Internet related scenarios [5, 6] To model client access patterns, we have followed an approach studied in [19] which uses state transition graphs to model client behaviors. Figure 6 depicts our model on how clients behave in the evaluation. A client moves between two states: SLEEP and STAY. A client moves from SLEEP state to STAY state by issuing a new request. During the STAY state, the client ....

D. Menasce, V. Almeida, R. Fonseca, and M. A. Mendes. Resource Management Policies for E-Commerce Servers. In Second Workshop on Internet Server Performance, May 1999.

.... news service) For client access patterns, we either use a real trace (in the case of on line forums) or develop a state transition graph to model client behaviors (in the case of auction and news service) The state transition graph approach was used in a recent study on e commerce user modeling [22]. 5.1 Online Forums Web site www.melissavirus.com provides an on line discussion forum on the notorious Melissa Virus after its tremendous destruction to computers worldwide. By observing how this Web site responds to client requests, we have developed an application to mimic the design and ....

....are issued one by one. 14 related to a topic (such as sports) changes, all pages which contain related headlines (e.g. URL class: cgi bin news topic=sports) are invalidated. To model client access patterns at a customized news site, we have followed the state transition graph approach used in [22]. Figure 7 shows our client behavior model. A client moves between two states: SLEEP and STAY. It moves from SLEEP state to STAY state by issuing a new request. During the STAY state, the client may automatically reload the same page every TRE seconds, or with a probability p, the client may ....

D. Menasce, V. Almeida, R. Fonseca, and M. A. Mendes. Resource Management Policies for E-Commerce Servers. In Second Workshop on Internet Server Performance, May 1999.

....data. E commerce workloads are composed of sessions. A session is a sequence of requests of different types made by a single customer during a single visit to a site [5] The allowed sequences of requests can be described by a state transition graph called Customer Behavior Model Graph (CBMG) [5, 12]. This graph has one node for each possible state (e.g. Home Page, Browse, Search, Select, Add, and Pay) and transitions between these states. A probability is assigned to each transition. Different types of users may be characterized by different CBMGs in terms of the transition probabilities. ....

D. A. Menasce, V. Almeida, R. Fonseca, and M. A. Mendes. Resource management policies for e-commerce servers. In Proc. Second Workshop on Internet Server Performance, Atlanta, GA, May 1st 1999. in conjunction with ACM SIGMETRICS 99/FCRC.

No context found.

D. A. Menasce, V. A. F. Almeida, R. Fonseca, and M. A. Mendes. Resource management policies for E-commerce servers. In Proc. ACM SIGMETRICS Workshop on Internet Server Performance, 1999.

No context found.

Mesnasce D, Almeida V, Fonesca R, Mendes M. Resource Management Policies for E-Commerce Servers.

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