MONTE : an implementation of an MPLS online traffic engineering tool
Resumen:
Multiservice networks require careful mapping of traffic in order to provide quality of service. Applying offline Traffic Engineering techniques leads to a better usage of resources and allows to assure some degree of quality of service. Even with those techniques applied, as network and traffic conditions change dynamically, the initial quality could be reduced.When addressing this problem, online Traffic Engineering has a major role. In MONTE project a solution for addressing this problem in Multiprotocol Label Switching networks was proposed and implemented in software. Such solution involves network discovering and monitoring, congestion detection, a corrective algorithm, and a mechanism for signalling changes in the network. The entire solution was conceived to work in real time and vendor independent. This paper explains the details of the solution and its implementation. Results validating the correct operation of the tool are also shown. This results were obtained through tests in a live network.
2009 | |
Multiprotocol label switching Online traffic engineering Load balance Performance monitoring Telecomunicaciones |
|
Inglés | |
Universidad de la República | |
COLIBRI | |
https://hdl.handle.net/20.500.12008/38635 | |
Acceso abierto | |
Licencia Creative Commons Atribución - No Comercial - Sin Derivadas (CC - By-NC-ND 4.0) |
Sumario: | Multiservice networks require careful mapping of traffic in order to provide quality of service. Applying offline Traffic Engineering techniques leads to a better usage of resources and allows to assure some degree of quality of service. Even with those techniques applied, as network and traffic conditions change dynamically, the initial quality could be reduced.When addressing this problem, online Traffic Engineering has a major role. In MONTE project a solution for addressing this problem in Multiprotocol Label Switching networks was proposed and implemented in software. Such solution involves network discovering and monitoring, congestion detection, a corrective algorithm, and a mechanism for signalling changes in the network. The entire solution was conceived to work in real time and vendor independent. This paper explains the details of the solution and its implementation. Results validating the correct operation of the tool are also shown. This results were obtained through tests in a live network. |
---|