| 1. |
- Carlsson, Patrik, et al.
(author)
-
Delay Performance in IP Routers
- 2004
-
Conference paper (peer-reviewed)abstract
- The main goals of the paper are towards an understanding of the delay process in best-effort Internet for both non-congested and congested networks. A dedicated measurement system is re-ported for delay measurements in IP routers, which follows specifications of the IETF RFC 2679. The system is using both passive measurements and active probing. Dedicated application-layer software is used to generate UDP traffic with TCP-like characteristics. Pareto traffic models are used to generate self-similar traffic in the link. The reported results are in the form of several impor-tant statistics regarding processing delay of a router, router delay for a single data flow, router delay for more data flows as well as end-to-end delay for a chain of routers. We confirm results reported earlier about the fact that the delay in IP routers is generally influenced by traffic characteristics, link conditions and, at some extent, details in hardware implementation and different IOS releases. The delay in IP routers usually shows heavy-tailed characteristics. It may also occasionally show extreme values, which are due to improper functioning of the routers.
|
|
| 2. |
- Constantinescu, Doru, et al.
(author)
-
Measurement of One-Way Internet Packet Delay
- 2004
-
Conference paper (peer-reviewed)abstract
- The paper reports on a dedicated measurement system for delay measurements in IP routers, which follows specifications of the IETF RFC 2679. The system uses both passive and active measurements. Dedicated application-layer software is used to generate traffic. Pareto traffic models are used to generate self-similar traffic in the link. Both packet inter-arrival times and packet sizes are matching real traffic models. A passive measurement system is used for data collection that is based on using several so-called Measurement Points, each of them equipped with DAG monitoring cards. Hashing is used for the identification and matching of packets. The combination of passive measurements and active probing, together with using the DAG monitoring system, gives us an unique possibility to perform precise traffic measurements as well as the flexibility needed to compensate for the lack of analytic solutions.
|
|
| 3. |
- Constantinescu, Doru, et al.
(author)
-
One-Way Transit Time Measurements
- 2004
-
Reports (peer-reviewed)abstract
- This report is a contribution towards a better understanding of traffic measurements associated with e2e delays occurring in best-effort networks. We describe problems and solutions associated with OWTT delay measurements, and give examples of such measurements. A dedicated measurement system is reported for delay measurements in IP routers, which follows specifications of the IETF RFC 2679. The system uses both passive measurements and active probing. Dedicated application-layer software is used to generate traffic. Pareto traffic models are used to generate self-similar traffic in the link. Both packet inter-arrival times and packet sizes are matching real traffic models. A passive measurement system is used for data collection that is based on using several so-called Measurement Points, each of them equipped with DAG monitoring cards. Hashing is used for the identification and matching of packets. The combination of passive and active measurements, together with using the DAG monitoring system, gives us an unique possibility to perform precise traffic measurements as well as the flexibility needed to compensate for the lack of analytic solutions. The real value of our study lies in the hop-by-hop instrumentation of the devices involved in the transfer of IP packets. The mixture of passive and active traffic measurements used, allows us to study changes in traffic patterns relative to specific reference points and to observe different contributing factors to the observed changes. This approach offers us the choice of better understanding diverse components that may impact on the performance of packet delay as well as to to measure queueing delays in operational routers.
|
|
| 4. |
- Erman, David, et al.
(author)
-
Measurement and Analysis of BitTorrent Signaling Traffic
- 2004
-
Conference paper (peer-reviewed)abstract
- BitTorrent is a second generation Peer-to-Peer application that has been recently developed as an alternative to the classical client-server model to reduce the load burden on content servers and networks. The protocol relies on the use of swarming techniques for distributing content. No search functionality is built into the protocol, and the signaling is geared only towards an efficient dissemination of data. The paper reports on measurement and analysis of BitTorrent traffic collected at the Blekinge Institute of Technology (BIT), Karlskrona, Sweden. We measure and analyze data from local BitTorrent client sessions at BIT. The characteristics of the signaling traffic exchanged among the participating peers in a BitTorrent distribution swarm are investigated. A dedicated approach based on combining instrumentation at the application layer with flow identification and extraction at the transport layer is used for traffic measurements.
|
|
| 5. |
- Ilie, Dragos, et al.
(author)
-
Measurement and Analysis of Gnutella Signaling Traffic
- 2004
-
Conference paper (peer-reviewed)abstract
- The paper reports on in-depth measurements and analysis of Gnutella signaling traffic collected at the Blekinge Institute of Technology (BIT), Karlskrona, Sweden. The measurements are based on a week-long packet trace collected with help of the well-known tcpdump application. Furthermore, a novel approach has been used to measure and analyze Gnutella signaling traffic. Associated with this, a dedicated tcptrace module has been developed and used to decode the packet trace, down to individual Gnutella messages. The measurement infrastructure consists of a Gnutella node running in ultrapeer mode and protocol decoding software. Detailed traffic characteristics have been collected and analyzed, such as session durations and interarrival times, and Gnutella message sizes and duration. Preliminary results show a high degree of variability of the Gnutella signaling traffic, which is mostly created by the QUERY messages. Furthermore, the Gnutella session interarrival times are observed to resemble the exponential distribution.
|
|
| 6. |
- Ilie, Dragos, et al.
(author)
-
Traffic Measurements of P2P Systems
- 2004
-
Conference paper (peer-reviewed)abstract
- The paper reports on a measurement infrastructure developed at the Blekinge Institute of Technology (BIT) with the purpose to do traffic measurements and analysis on Peer-to-Peer (P2P) traffic. The measurement methodology is based on using application logging as well as link-layer packet capture. This offers the possibility to measure application layer information with link-layer accuracy. Details are reported on this methodology, together with description of the BIT measurement infrastructure. The paper also reports on traffic measurements done on BitTorrent and Gnutella protocols from an end-client perspective, together with some measurement results of salient protocol characteristics. Preliminary results show a high degree of variability of the BitTorrent and Gnutella traffic, where in the case of Gnutella a large contribution is given by the signaling traffic.
|
|
| 7. |
- Jena, Ajit K., et al.
(author)
-
Application Layer Tracing for Performance Evaluation
- 2002
-
Conference paper (peer-reviewed)abstract
- The paper reports on a set of non-intrusive tools that can be used for the performance evaluation of "classical" Internet applications such as Email, WWW, and FTP. At the application layer, key factors influencing the events are stochastic aspects of user behavior, protocol characteristics and sizes of contents transferred. A multi pronged approach involving application log analysis, direct probe of the WWW pages, and network flow analysis has been used. The measurement infrastructure is presented and diverse software tools for gathering of application layer properties are described in detail. The motivation is to build stochastic models for the key elements of application layer that can be further used to build up an integrated end-to-end performance testbed for conducting traffic engineering experiments.
|
|
| 8. |
- Jena, Ajit K., et al.
(author)
-
End-User Performance of WWW Services
- 2001
-
Conference paper (peer-reviewed)abstract
- The global Internet has seen tremendous growth in terms of nodes and user base as well as of types of applications. One of the most important consequences of this growth is related to an increased complexity of the traffic models experienced in the networks. Each application has a set of unique characteristics in terms of the way it performs its transactions as well as the way transaction processing profile maps onto unique network resource requirements. In order to support Internet applications effectively, it is therefore important to understand and to characterize the application level transactions and also to investigate their scaling properties. Recent advances in high resolution traffic monitoring and analyzing capabilities have enabled us to build up realistic models for the TCP/IP protocol stack with diverse network applications. In this paper we report investigations of classical applications such as FTP, SMTP, and HTTP to evaluate end-to-end performance requirements and accordingly to assess end-user performance like Service Level Agreement (SLA) for WWW. Our results show the presence of a robust correlation structure in the traffic streams that has a fundamental bearing on the user perceived quality of the applications.
|
|
| 9. |
- Jena, Ajit K., et al.
(author)
-
Modeling and Analysis of HTTP Traffic
- 2000
-
Conference paper (peer-reviewed)abstract
- The global Internet has seen tremendous growth in terms of nodes and user base as well as of types of applications. One of the most important consequences of this growth is related to an increased complexity of the traffic experienced in these networks. Each application has a set of unique characteristics in terms of performance characteristics, transactions as well as the way the transaction processing profile maps onto unique network resource requirements. In order to support Internet applications effectively,it is therefore important to understand and to characterize the application level transactions as well as the effect of different TCP/IP control mechanisms on application-level parameters. It is the goal of this paper to model and to evaluate the characteristics of World Wide Web traffic. Results are reported on measuring, modeling and analysis of specific Hyper Text Transfer Protocol traffic collected from different (classes of) sites together with methodologies used for capturing HTTP flows as well as for modeling. The paper concludes with a discussion on the structure of Web pages and a model for the generation of the number of embedded pages in a Web page is suggested.
|
|
| 10. |
- Jena, Ajit K., et al.
(author)
-
Modeling and Evaluation of Internet Applications
- 2002
-
Reports (other academic/artistic)abstract
- The paper presents a modeling and evaluation study of the characteristics of several "classical" Internet applications (SMTP, HTTP and FTP) in terms of user behavior, nature of contents transferred and application layer protocol exchanges. Results are reported on measuring, modeling and analysis of application layer traces collected, at both the client and the server end, from different environments such as university networks and commercial Frame Relay networks. The methodologies used for capturing traffic flows as well as for modeling are reported. Statistical models have been developed for diverse parameters of applications (e.g., HTTP document sizes, FTP file sizes, and SMTP message sizes), which can be useful for building synthetic workloads for simulation and benchmarking purposes. All three applications possess a session oriented structure. Within each session, a number of transactions are performed. For the above mentioned applications, the number of transactions that may occur during a session has been also modeled.
|
|
