Optimizing the Performance of the Iptables Stateful NAT44 Solution

as seious limitations. Our Internet Draft has roposed a Abstract-The stteful NAT44 performance of iptables is a rformance measurements of iptables published in bnchmarking methodology for tteful NATxy (x y re in reseach important issue papers when do no it is used comply as wi a h staeful the NAT44 requirements gateway of of a CGN (Carrier-Gra NAT) sytem. The prfomance RFC 2544 and RFC4814 nd th usabili y f their results {4 6 ) gateways and made t possble to p rform the classic RFC 2544 easurement procedures like thughput, meauements of iptables publishd in researc paers do not has serious limitations Our Internet Draft has proposed a l tency, f me l rate, t with t t ful NATxy g t way comply with the requirements of RFC 2544 and RFC 4814 benchmarking methodology for stateful NATxy x, y are in and the usability of their results has serious limitations. Our using RFC 4814 pseudorandom port numbers. I has also {4, 6}) gateways and made it possible to perform the classic Internet Draft has proposed a benchmarking methodology for defined new performanc metcs specific to staeful testng stateful NATxy (x, y are in {4, 6}) gateways and made it possible RFC 2544 me u em nt pr ed lik th o ghput, quantify the connection setup and connec ion tear down to perform the classic RFC 2544 measurement procedures like latency, frame loss rate, etc. with stateful NATxy gatewayspeformance of stateful NATxy gatways. In our current throughput, latency, frame loss rate, etc. with stateful NATxy using RFC 4814 pseudorandom p rt numb rs. It has also pper, gateways we using ex RFC mine 4814 how the pseudorandoperform nce port of numbrs. iptables It defined new pe formance metrics specific to stateful test ng has also defined new performance metrics specific to stateful depends on varius settigs, and also if certain traeoffs to quanify the connection setup and connection ear downexist We msure maximum connection establishm testing to quantify the connection seup and connection tear performance of stateful NATxy gateways. In our currentdown performance of stateful NATxy gatewys. In our current pper, we exame how the performance of iptable rate throughput and tear down rate of iptables as well as its memory consumption a a functin of hash table size par, we examine how the performance of iptables depends on depends on various settings and also if certain tradeoffs various settings, and also if certain tradeoffs exist. We measure alw y u i g 40 mlli conn ti . We di cl e ll exist We measure the maximum conn ection establishment the maximum connection establishment rate, throughput and measurement details and results. We recommend new r t , th ghput nd t d w rat f ipt bl a well tear down rate of iptables as well as its memory consumption as a function of hash table size always using 40 million connectons. settings that enable network operators to achiev its memory con umption a a function of hash table size signifiantly higher prfrmance than using th traditiona We disclose all measurement details and results. We recommend always usng 40 million connections. We disclose all newones settings that enable ntwork operators to achive mea uremen deta ls and results. We recomme d new significantly higher performance than using the traditional ones. settings that enable network operators to achieve