Estudi d'implantació de la IPv6 al CTTC
Tutor / director / evaluatorMeseguer Pallarès, Roc
Document typeBachelor thesis
Rights accessOpen Access
In recent years, the number of devices connected to Internet has increased exponentially and in fact, it is estimated that the demand keep on increasing. When IPv4 protocol was introduced, this evolution was not expected. Over the years several limitations have arosen, including scalability and the unsustainable growth of the Internet routing table. In the late 90s, IPv6 protocol was defined to solve these issues and add new features. Currently, the Centre Tecnològic de Telecomunicacions de Catalunya (CTTC) has been forced to propose IPv6 deployment on its network mainly caused by the IPv4 address depletion, the increasing demand for devices connectivity and the need to support IPv6. RedIris and CSUC are encouraging and supporting the affiliated institutions like CTTC to deploy IPv6. This project is carried out by the IT department of CTTC, the Centre de Serveis Informàtics (CSI). The purpose of this project is, on the one hand, to analyse and acknowlege the main features of IPv6 protocol and show the differences with IPv4. It also aims to implement native IPv6 network in CTTC. On the other hand, the CTTC needs to give support with IPv6 in some basic services like DNS and web services. Finally, it is thought to compare IPv6 and IPv4 performance on the CTTC network. We analysed the network equipment and the network topology before proposing an IPv6 compatible solution. This causes the replacement of some network equipment. To gradually integrate IPv6 in the CTTC network we have chosen the dual-stack deployment. IPv6 is deployed in parallel with IPv4 on all infrastructure and services. We are adapting the DNS and the website service to IPv6 requests. In this project we are measuring the performance of IPv6 versus IPv4. We are running some tests to evaluate the performance through the calculation of the Round-Trip Time (RTT), throughput and Time To First Byte (TTFB) parameters. The results of the performance tests in the CTTC network are not showing important differences between the two protocols. The RTT is larger in IPv6, the throughput is higher in IPv4 and the TTFB depends on if the dual stack is enabled. The network latency is lower in IPv6 on a dual stack scenario. However, when there is only one protocol set the result is opposite. Thedistance between the origin and the destination of communications is relevant because this affects the latency of the connection. This is due to the number of hops to destination, the network type and the service provided. In another test, we are comparing IPv6 with NAT (Network Address Translation) to evaluate this alternative method as a solution to the IPv4 address depletion. The results obtained in the comparison are quite similar.