nico/master-thesis
1
0
Fork 0
Code Issues Pull requests Projects Releases Wiki Activity

sed -i -e 's/jool/Jool/g' -e 's/tayga/Tayga/g' *.tex

Browse source 
+ manual fixes on that one
This commit is contained in:
Nico Schottelius 2019-08-22 12:09:05 +02:00
parent c2e8b9bf75
commit 971c0e02c3
4 changed files with 20 additions and 20 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

12
doc/Design.tex
View file

@ -270,7 +270,7 @@ to solve this problem:
and inserts the original packet afterwards back into the and inserts the original packet afterwards back into the
switch. Figure \ref{fig:p4switchstateful} shows the flow of a packet switch. Figure \ref{fig:p4switchstateful} shows the flow of a packet
with stateful translation in detail. with stateful translation in detail.
\item With tayga we rely on the Linux kernel NAT44 capabilities \item With Tayga we rely on the Linux kernel NAT44 capabilities
\item Jool implements its own stateful mechanism based on port \item Jool implements its own stateful mechanism based on port
ranges ranges
\end{itemize} \end{itemize}
@ -296,7 +296,7 @@ commands allow interactive testing on TCP and UDP connections, while
the iperf commands fully utilise the available bandwidth with test the iperf commands fully utilise the available bandwidth with test
data. data.
The socat and iperf commands are used to verify all three NAT64 The socat and iperf commands are used to verify all three NAT64
implementations (p4, tayga, jool). implementations (p4, Tayga, Jool).
\begin{table}[htbp] \begin{table}[htbp]
\begin{center}\begin{minipage}{\textwidth} \begin{center}\begin{minipage}{\textwidth}
\begin{tabular}{| c | c | c |} \begin{tabular}{| c | c | c |}
@ -377,9 +377,9 @@ networks and IPv6 addresses are used:
\hline \hline
2001:db8:42::a00:2a & In-network IPv6 address mapped to 10.0.0.42 (p4)\\ 2001:db8:42::a00:2a & In-network IPv6 address mapped to 10.0.0.42 (p4)\\
\hline \hline
2001:db8:23::a00:2a & IPv6 address mapped to 10.0.0.42 (tayga) \\ 2001:db8:23::a00:2a & IPv6 address mapped to 10.0.0.42 (Tayga) \\
\hline \hline
2001:db8:23::2a & IPv6 address mapped to 10.0.0.42 (jool)\\ 2001:db8:23::2a & IPv6 address mapped to 10.0.0.42 (Jool)\\
\hline \hline
\end{tabular} \end{tabular}
\end{minipage} \end{minipage}
@ -405,9 +405,9 @@ from the 10.0.0.0/8 range as follows:
\hline \hline
10.0.0.66 & In-network IPv4 address mapped to 2001:db8:42::42 (p4)\\ 10.0.0.66 & In-network IPv4 address mapped to 2001:db8:42::42 (p4)\\
\hline \hline
10.0.1.42 & IPv4 address mapped to 2001:db8:42::42 (tayga)\\ 10.0.1.42 & IPv4 address mapped to 2001:db8:42::42 (Tayga)\\
\hline \hline
10.0.1.66 & IPv4 address mapped to 2001:db8:42::42 (jool)\\ 10.0.1.66 & IPv4 address mapped to 2001:db8:42::42 (Jool)\\
\hline \hline
\end{tabular} \end{tabular}
\end{minipage} \end{minipage}

2
doc/Introduction.tex
View file

@ -102,6 +102,6 @@ software~\cite{_implem_your_switc_target_with_bmv2}. Figure
\ref{fig:v6v4mixed} shows how the design differs for an in-network \ref{fig:v6v4mixed} shows how the design differs for an in-network
solution. solution.
Even on fast CPUs, software solutions like Even on fast CPUs, software solutions like
tayga~\cite{lutchansky:_tayga_simpl_nat64_linux} Tayga~\cite{lutchansky:_Tayga_simpl_nat64_linux}
can be CPU bound (see section \ref{results:softwarenat64}) and are can be CPU bound (see section \ref{results:softwarenat64}) and are
incapable of translating protocols at line speed. incapable of translating protocols at line speed.

14
doc/Results.tex
View file

@ -19,8 +19,8 @@ NAT64~\cite{schottelius:thesisrepo}. It contains parsers
for all related protocols (IPv6, IPv4, UDP, TCP, ICMP, ICMP6, NDP, for all related protocols (IPv6, IPv4, UDP, TCP, ICMP, ICMP6, NDP,
ARP), supports EAMT as defined by RFC7757 ~\cite{rfc7757}, and is ARP), supports EAMT as defined by RFC7757 ~\cite{rfc7757}, and is
feature equivalent to the two compared software solutions feature equivalent to the two compared software solutions
tayga~\cite{lutchansky:_tayga_simpl_nat64_linux} and Tayga~\cite{lutchansky:_Tayga_simpl_nat64_linux} and
jool~\cite{mexico:_jool_open_sourc_siit_nat64_linux}. Jool~\cite{mexico:_Jool_open_sourc_siit_nat64_linux}.
Due to limitations in the P4 environment of the Due to limitations in the P4 environment of the
NetFPGA environment, the BMV2 implementation NetFPGA environment, the BMV2 implementation
is more feature rich. is more feature rich.
@ -429,7 +429,7 @@ warm up phase.\footnote{iperf -O 10 parameter, see section \ref{design:tests}.}
% ok % ok
% ---------------------------------------------------------------------- % ----------------------------------------------------------------------
\subsection{\label{results:benchmark:summary}Benchmark Summary} \subsection{\label{results:benchmark:summary}Benchmark Summary}
Overall \textbf{tayga} has shown to be the slowest translator with an achieved Overall \textbf{Tayga} has shown to be the slowest translator with an achieved
bandwidth of \textbf{about 3 Gbit/s}, followed by \textbf{Jool} that translates at bandwidth of \textbf{about 3 Gbit/s}, followed by \textbf{Jool} that translates at
about \textbf{8 Gbit/s}. \textbf{Our solution} is the fastest with an almost line rate about \textbf{8 Gbit/s}. \textbf{Our solution} is the fastest with an almost line rate
translation speed of about \textbf{9 Gbit/s}. translation speed of about \textbf{9 Gbit/s}.
@ -443,7 +443,7 @@ in the UDP benchmarks incorporates the packets loss (compare tables
\ref{tab:benchmarkv6v4udp} and \ref{tab:benchmarkv6v4udp}). \ref{tab:benchmarkv6v4udp} and \ref{tab:benchmarkv6v4udp}).
Both software solutions showed significant loss of packets in the UDP Both software solutions showed significant loss of packets in the UDP
based benchmarks (tayga: up to 91\%, jool up to 71\%), while the based benchmarks (Tayga: up to 91\%, Jool up to 71\%), while the
P4/NetFPGA showed a maximum of 0.01\% packet loss. Packet loss is only P4/NetFPGA showed a maximum of 0.01\% packet loss. Packet loss is only
recorded by iperf for UDP based benchmarks, as TCP packets are confirmed and recorded by iperf for UDP based benchmarks, as TCP packets are confirmed and
resent if necessary. resent if necessary.
@ -455,7 +455,7 @@ identical in different benchmark runs.
The CPU load for TCP based benchmarks with Jool was almost negligible, The CPU load for TCP based benchmarks with Jool was almost negligible,
however for UDP based benchmarks one core was almost 100\% however for UDP based benchmarks one core was almost 100\%
utilised. In all benchmarks with tayga, one CPU was fully utilised. In all benchmarks with Tayga, one CPU was fully
utilised. When the translation for P4/NetFPGA happens within the utilised. When the translation for P4/NetFPGA happens within the
NetFPGA card, there was no CPU utilisation visible on the NAT64 host. NetFPGA card, there was no CPU utilisation visible on the NAT64 host.
@ -476,11 +476,11 @@ connections might be affected by the load generator limits. While
there is no visible evidence in our results, this problem might become there is no visible evidence in our results, this problem might become
more significant with higher speed links. more significant with higher speed links.
While tayga's performance is reduced with the growing number of While Tayga's performance is reduced with the growing number of
parallel connections, both Jool and our P4/NetFPGA implementations parallel connections, both Jool and our P4/NetFPGA implementations
vary only slighty. vary only slighty.
Overall the performance of tayga, a Linux user space program, is as Overall the performance of Tayga, a Linux user space program, is as
expected. We were surprised about the good performance of Jool, which, expected. We were surprised about the good performance of Jool, which,
while slower than the P4/NetFPGA solution, is almost on par with our solution. while slower than the P4/NetFPGA solution, is almost on par with our solution.
% ---------------------------------------------------------------------- % ----------------------------------------------------------------------

12
doc/appendix.tex
View file

@ -1300,7 +1300,7 @@ root@ESPRIMO-P956:~#
\end{verbatim} \end{verbatim}
%--------------------------------------------------------------------------------------------------------- %---------------------------------------------------------------------------------------------------------
\section{\label{benchmark:tayga}Tayga} \section{\label{benchmark:Tayga}Tayga}
Tayga is installed from the regular package database: Tayga is installed from the regular package database:
\begin{verbatim} \begin{verbatim}
ii tayga 0.9.2-6 amd64 userspace stateless NAT64 ii tayga 0.9.2-6 amd64 userspace stateless NAT64
@ -1366,7 +1366,7 @@ net.ipv6.conf.all.forwarding = 1
[15:20] nsg-System:~# sysctl -w net.ipv4.ip_forward=1 [15:20] nsg-System:~# sysctl -w net.ipv4.ip_forward=1
net.ipv4.ip_forward = 1 net.ipv4.ip_forward = 1
\end{verbatim} \end{verbatim}
And we test NAT64 with tayga: And we test NAT64 with Tayga:
\begin{tiny} \begin{tiny}
\begin{verbatim} \begin{verbatim}
nico@ESPRIMO-P956:~$ ping -c2 10.0.1.42 nico@ESPRIMO-P956:~$ ping -c2 10.0.1.42
@ -1407,9 +1407,9 @@ rtt min/avg/max/mdev = 0.240/0.320/0.400/0.080 ms
nico@ESPRIMO-P956:~$ nico@ESPRIMO-P956:~$
\end{verbatim} \end{verbatim}
%--------------------------------------------------------------------------------------------------------- %---------------------------------------------------------------------------------------------------------
\section{\label{benchmark:jool}Jool} \section{\label{benchmark:Jool}Jool}
We install Jool 4.0.1 from source from We install Jool 4.0.1 from source from
\url{https://www.jool.mx/en/download.html} as follows: \url{https://www.Jool.mx/en/download.html} as follows:
\begin{tiny} \begin{tiny}
\begin{verbatim} \begin{verbatim}
nico@nsg-System:~$ wget https://github.com/NICMx/Jool/releases/download/v4.0.1/jool_4.0.1.tar.gz nico@nsg-System:~$ wget https://github.com/NICMx/Jool/releases/download/v4.0.1/jool_4.0.1.tar.gz
@ -1426,8 +1426,8 @@ We enable forwarding:
sysctl -w net.ipv4.conf.all.forwarding=1 sysctl -w net.ipv4.conf.all.forwarding=1
sysctl -w net.ipv6.conf.all.forwarding=1 sysctl -w net.ipv6.conf.all.forwarding=1
\end{verbatim} \end{verbatim}
We configure jool to map the network prefixes and setup iptables to We configure Jool to map the network prefixes and setup iptables to
redirect the traffic into the jool instance: redirect the traffic into the Jool instance:
\begin{tiny} \begin{tiny}
\begin{verbatim} \begin{verbatim}
[16:53] nsg-System:~# modprobe jool_siit [16:53] nsg-System:~# modprobe jool_siit