\subsection{Preflight checklist}
-
Before actual certification is commenced we perform a manual test using a
checklist.
If any of the checks fail we immediately reject the product.
+The checklist is given in the table below. All commands in \texttt{monospace}
+are to be run in a terminal. Commands prefixed with a \texttt{\#} should be run
+with root permissions. Commands prefixed with a \texttt{\$} should be run with
+user permissions.
-\begin{enumerate}
- \item Is the product complete?
- \item Does the product come with a manual or quick start guide?
- \item Is it possible to get the product in a usable state?
- \item Can we use the product to initiate a connection in a corruption
- free environment?
-\end{enumerate}
-
-\textbf{DIT MOET VERANDERD WORDEN, HET PRODUCT ZOU AFGEKEURD WORDEN!}
+\begin{longtable}{|l|rp{.8\linewidth}|}
+ \hline
+ Check 1 & \multicolumn{2}{l|}{Get the SUT in a workable state.}\\
+ \hline
+ \multirow{3}{*}{Course of action}
+ & 1. & Import the VirtualBox image into VirtualBox.\\
+ & 2. & Boot the vm.\\
+ & 3. & Verify the SUT booted successfully and the network modules are
+ loaded.\\
+ \hline
+ Passed & \multicolumn{2}{l|}{\textit{Yes/No}}\\
+ \hline\hline
+ Check 2 & \multicolumn{2}{l|}{Verify the SUT is complete.}\\
+ \hline
+ \multirow{5}{*}{Course of action}
+ & 1. & Boot the SUT as in \emph{Check 1}.\\
+ & 2. & Verify the loopback device exists by running
+ \texttt{\$ ifconfig}.\\
+ & 3. & Verify the \emph{echo-server} is present on the system by running
+ \texttt{\$ file code/server/Main.java}\\
+ & 4. & Verify \emph{Scapy} is present on the system by running
+ \texttt{\$ scapy}.\\
+ & 5. & Verify all scripts used for testing are present on the system.\\
+ \hline
+ Passed & \multicolumn{2}{l|}{\textit{Yes/No}}\\
+ \hline\hline
+ Check 3 & \multicolumn{2}{l|}{Initialize the testing environment..}\\
+ \hline
+ \multirow{6}{*}{Course of action}
+ & 1. & Boot the SUT as in \emph{Check 1}.\\
+ & 2. & Setup iptables by executing
+ \texttt{\# code/iptables.sh}~\footnote{The IPTables script ensures
+ that the OS does not drop packets due to an the unknown source.}\\
+ & 3. & Navigate to the working directory by running
+ \texttt{\$ cd /home/student/tt2015}\\
+ & 4. & Compile the echo server by running
+ \texttt{\# cd code/server \&\& make}\\
+ & 5. & Start the echo server by running
+ \texttt{\# cd code/server \&\& java Main}\\
+ & 6. & Generate all test cases by running
+ \texttt{\$ python code/client/gen.py}\\
+ \hline
+ Passed & \multicolumn{2}{l|}{\textit{Yes/No}}\\
+ \hline\hline
+ Check 4 & \multicolumn{2}{l|}{Test the tool environment.}\\
+ \hline
+ \multirow{3}{*}{Course of action}
+ & 1. & Initialize the SUT as in \emph{Check 3}\\
+ & 2. & Execute the test script by running
+ \texttt{\# code/client/helloworld.py}\\
+ & 3. & Verify the console displays a success message.\\
+ \hline
+ Passed & \multicolumn{2}{l|}{\textit{Yes/No}}\\
+ \hline\hline
+ Check 5 & \multicolumn{2}{l|}{All test inputs and scripts are present.}\\
+ \hline
+ \multirow{2}{*}{Course of action}
+ & 1. & Boot the SUT as in \emph{Check 1}.\\
+ & 2. & Verify that the test generation script is present by running
+ \texttt{\$ file code/client/test.py}\\
+ \hline
+ Passed & \multicolumn{2}{l|}{\textit{Yes/No}}\\
+ \hline
+ \caption{Preflight checklist\label{tbl:preflight}}
+\end{longtable}
\subsection{Testing of SUT}
-
-The SUT is a series of services for other computer programs with no end-user
-facing
-interface. Therefore the SUT will be tested solely by calling it's services
-through various automated scripts. An automated test suite will be available
-which executes all these automated scripts and aggregates their results to
-asses whether or not the SUT has passed the test.
+The SUT is a series of services for other computer programs with no end-user
+facing interface. Therefore the SUT will be tested solely by calling it's
+services through various automated scripts. An automated test suite will be
+available which executes all these automated scripts and aggregates their
+results to asses whether or not the SUT has passed the test.
The implementation of the SUT is tested using black box testing techniques. A
series of tests asses the correctness of the implementation with regards to the
TCP specification. These tests are specified in Table~\textbf{referentie naar
-tests-tabel}. The test cases aim to cover the most interesting parts of the
-TCP specification.
+tests-tabel}. The test cases aim to cover the most interesting parts of the TCP
+specification.
To cover the TCP specification as complete as possible while still maintaining
a feasible test suite the tests are divided into equivalence partitions. Below
these partitions are given.
\begin{enumerate}
- \item Single valid request.
- \item Single invalid request with.
- \begin{enumerate}
- \item Corrupted checksum.
- \item Corrupted payload with.
+ \item \emph{Number of segments} in request~\footnote{A request is
+ considered establishing a connection (handshake) and a number of
+ payload segments}
+ \begin{enumerate}
+ \item 0 payload segments
+ \item 1 payload segments
+ \item n=small payload segments (1 byte)
+ \item n=big payload segments (65495 bytes)
+ \end{enumerate}
+ \item \emph{source port}
+ \begin{enumerate}
+ \item Correct
+ \item Incorrect
+ \end{enumerate}
+ \item \emph{destination port}
+ \begin{enumerate}
+ \item Correct
+ \item Incorrect
+ \end{enumerate}
+ \item Bit errors in \emph{payload}
+ \begin{enumerate}
+ \item Correct payload
+ \item Payload with bit flips that do not show in checksum
+ \item Payload with bit flips that do show in checksum
+ \end{enumerate}
+ \item \emph{checksum}
\begin{enumerate}
- \item Oven number of bits flipped.
- \item Odd number of bits flipped.
+ \item Correct
+ \item Incorrect
\end{enumerate}
- \item Corrupted source address.
- \item Corrupted destination address.
- \end{enumerate}
- \item Multiple valid requests received in order.
- \item Multiple valid requests received out of order.
- \item Mixed valid and invalid requests with.
- \begin{enumerate}
- \item Missing packets.
- \item Corrupted checksum.
- \item Corrupted payload with.
+ \item \emph{Segment order}
\begin{enumerate}
- \item Oven number of bits flipped.
- \item Odd number of bits flipped.
+ \item Correct
+ \item Out of order
+ \item Missing Segments
\end{enumerate}
- \item Corrupted source address.
- \item Corrupted destination address.
- \end{enumerate}
\end{enumerate}
-Above enumeration results in a total of $14$ test cases.
+These partitions were chosen since they correspond to key parts of the TCP
+specification. A single request consists of a number of packets that sent a
+to the \emph{echo-server} and back. The TCP specification state that such a
+transaction requires the following messages.
+
+\begin{center}
+ Script $-$ SYN $A$ $\rightarrow$ SUT \\
+ Script $\leftarrow$ SYN-ACK $(A+1)$ $B$ $-$ SUT \\
+ Script $-$ ACK $(A+1)$ $(B+1)$ $\rightarrow$ SUT \\
+ Script $-$ ACK-PUSH $(A+1)$ $(B+2)$ \emph{msg} $\rightarrow$ SUT \\
+ Script $\leftarrow$ ACK $(A+2+msg_{length})$ $(B+3)$ $-$ SUT \\
+ Script $\leftarrow$ ACK-PUSH $(A+2+msg_{length})$ $(B+3)$ \emph{msg} $-$ SUT
+\end{center}
+
+%
+% één na laatste packket, moet dit B+3 of B+2 zijn?
+%
+
+TCP segments are send over a TCP connection from a \emph{source} to a \emph{destination port}. Therefore segments which are received that have a
+source or destination port set to an incorrect value should not be regarded
+as segments belonging to the connection by the SUT.
-\vspace{3mm}
-\textbf{Hier daadwerkelijke test cases tabel}
-\vspace{3mm}
+TCP uses a \emph{checksum} to catch any error introduced in headers, when this
+checksum does not match the actual computed checksum the SUT should
+disregard the received segment.
-To reduce the number of tests that need to be executed when testing the SUT and
-to construct a feasible test suite these tests are combined according
-to Table~(\textbf{referentie naar decision table}).
+The TCP checksum is also an inherently weak one, as it is simply the
+bitwise negation of the addition, in ones complement arithmetic,
+of all 16 bit words in the header and data of the segment (excluding the
+checksum itself). Therefore any \emph{bit error} where the ones complement value
+of one word
+increases by one, and the value of another decreases by one, is undetected.
+The SUT should exhibit the same behavior and accept packets where these type
+of bit errors occur.
-\vspace{3mm}
-\textbf{Hier daadwerkelijke decision tabel}
-\vspace{3mm}
+TCP guarantees that segments are delivered \emph{in order}
+,even when they are received
+out of order and that missing segments are resend. The SUT should
+exhibit the same behavior. If segments are received out of order it should
+either reassemble them when the missing packet has arrived or request them to
+be resend when the Missing segments should be re-requested (by ACK-ing
+the correct sequence number).
+
+\bigskip
+
+Partitions 2 to 6 are tested using pairwise testing to keep the number of test
+cases feasible. The pairs are then all *except some where it does not make sense
+to do so) tested with the different request sizes of partition 1.
+
+This is expressed in Table~\ref{table:testpairs}. In this table the first five
+columns represent the different options for the partitions 2 to 6 of the above
+enumeration. The last four columns are the different number segments as
+described in the partition 1 of the above enumeration. These cells identify
+individual test cases by a number. An \xmark in the cell indicates that this
+test case can not be created as it is not possible with that number of segments
+(eg. sending segments out of order when the number of segments is 1).
+
+\newcounter{TCC}
+\setcounter{TCC}{1}
+\newcommand{\doTCC}{\theTCC \stepcounter{TCC}}
+\begin{table}[H]
+ \centering
+ \begin{tabular}{|l|l|l|l|l|l||l|l|l|l|}
+ \hline
+ & \multicolumn{9}{c|}{\textbf{Partition}}\\
+ \hline
+ & 4 & 5 & 3 & 6 & 2 & 1a & 1b & 1c & 1d\\
+ \hline\hline
+ \multirow{9}{*}{Instance}
+ & a & a & a & a & a & \doTCC & \doTCC & \doTCC & \doTCC\\
+ & a & b & b & c & b & \xmark & \xmark & \doTCC & \doTCC\\
+ & c & a & b & a & b & \xmark & \xmark & \doTCC & \doTCC\\
+ & c & b & a & c & a & \xmark & \xmark & \doTCC & \doTCC\\
+ & b & a & b & c & a & \xmark & \xmark & \doTCC & \doTCC\\
+ & b & b & a & b & b & \xmark & \xmark & \doTCC & \doTCC\\
+ & c & b & b & a & b & \xmark & \doTCC & \doTCC & \doTCC\\
+ & b & b & b & a & b & \xmark & \doTCC & \doTCC & \doTCC\\
+ & a & b & b & b & a & \xmark & \xmark & \doTCC & \doTCC\\
+ \hline
+\end{tabular}
+\caption{Combinations of test cases}
+\label{table:testpairs}
+\end{table}
\subsection{Quality, completeness and coverage of tests}
-\subsubsection{Quality}
The network packets used in testing are constructed from prerecorded, known to
be correct, network traffic. These packets are then modified with well used and
field tested tools. Due to this the chance of errors in the test cases is quite
there is a fault in the SUT and can not result directly in the conclusion that
there actually is one.
-\subsubsection{Coverage}
+\bigskip
+
Due to the nature of black-box testing coverage of the code in the
implementation of the SUT is unknown. However completeness of the tests over
the specification of the SUT can be assessed.
-\subsubsection{Completeness}
+\bigskip
+
Due to the clear and exhaustive specification of TCP the completeness of the
test suite can be clearly assessed.
As always, $100\%$ completeness is not feasible, therefore test cases are
carefully selected to cover the most interesting parts of the TCP specification
-to ensure a test suite which covers at least \textbf{percentage}\% of the
-specification.
+to ensure a complete but feasible test suite.
+
+To further increase the coverage of the test suites tests are randomized. The
+tests which test the handling of \emph{bit errors}, changes in the \emph{packet
+order} and \emph{dropped packets} randomize where they introduce an error. The
+test suite runs these tests multiple times to increase the likelihood that they
+discover a fault which is only present when an error occurs in a certain
+position.
To further decrease the number of tests needed test cases are divided into
equivalence partitions and the combination of cases as described in
-Table~(\textbf{referentie naar decision table}) ensures that all partitions are
+Table~\ref{table:testpairs} ensures that all partitions are
covered and the number of individual tests is still feasible.
+
+%
+% wat ik ook probeer ik krijg de eerste collum
+% zijn tekst niet verticaal gecentered
+%
+
+
+\subsection{Test cases}
+
+Before every test case use the following steps to initialize the testing environment.
+
+\begin{enumerate}
+ \item Boot the vm using VirtualBox.
+ \item Setup iptables by executing \texttt{\# code/iptables.sh}
+ \item Navigate to the working directory by running \texttt{\$ cd /home/student/tt2015}
+ \item Start the echo server by running \texttt{\# cd code/server \&\& java Main}
+\end{enumerate}
+
+\begin{longtable}{|p{.2\linewidth}|p{.8\linewidth}|}
+ \hline
+ Nr & 1 \\\hline
+ Title & Single valid request with 1byte payload. \\\hline
+ Input & Generated packets. \\\hline
+ Expected output & Packets echoed back by Echo-Server. \\\hline
+ \multirow{2}{*}{Course of action}
+ & 1. Use the steps listed above in order to start the SUT. \\
+ & 2. Execute the script by running \texttt{\# code/client/tests/1.py} \\\hline
+ Valid trace & Verify that the script prints 'Success'. \\\hline
+ \hline
+
+ Nr & 2 \\\hline
+ Title & Single valid request with 65495bytes payload. \\\hline
+ Input & Generated packets. \\\hline
+ Expected output & Packets echoed back by Echo-Server. \\\hline
+ \multirow{2}{*}{Course of action}
+ & 1. Use the steps listed above in order to start the SUT. \\
+ & 2. Execute the script by running \texttt{\# code/client/tests/2.py} \\\hline
+ Valid trace & Verify that the script prints 'Success'. \\\hline
+ \hline
+
+ Nr & 3 \\\hline
+ Title & 5 valid requests with 1byte payload. \\\hline
+ Input & Generated packets. \\\hline
+ Expected output & Packets echoed back by Echo-Server, in the same order as the client sent them. \\\hline
+ \multirow{2}{*}{Course of action}
+ & 1. Use the steps listed above in order to start the SUT. \\
+ & 2. Execute the script by running \texttt{\# code/client/tests/3.py} \\\hline
+ Valid trace & Verify that the script prints 'Success'. \\\hline
+ \hline
+
+ Nr & 4 \\\hline
+ Title & 5 valid requests with 65495bytes payload. \\\hline
+ Input & Generated packets with 65495bytes payload. \\\hline
+ Expected output & Packets echoed back by Echo-Server, in the same order as the client sent them. \\\hline
+ \multirow{2}{*}{Course of action}
+ & 1. Use the steps listed above in order to start the SUT. \\
+ & 2. Execute the script by running \texttt{\# code/client/tests/4.py} \\\hline
+ Valid trace & Verify that the script prints 'Success'. \\\hline
+ \hline
+
+ Nr & 5 \\\hline
+ Title & 5 valid requests with 1byte payload sent out of order. \\\hline
+ Input & Generated packets with 1byte payload, two packets are swapped in position. \\\hline
+ Expected output & All requests sent up to and including
+ the swapped packet with the lowest sequence number, the remaining packets are dropped. \\\hline
+ \multirow{2}{*}{Course of action}
+ & 1. Use the steps listed above in order to start the SUT. \\
+ & 2. Execute the script by running \texttt{\# code/client/tests/5.py} \\\hline
+ Valid trace & Verify that the script prints 'Success'. \\\hline
+ \hline
+
+ Nr & 6 \\\hline
+ Title & Request with corrupted source port. \\\hline
+ Input & Generated packets with 1byte payload, in these packets the source port number is increased by one. \\\hline
+ Expected output & - \\\hline
+ \multirow{2}{*}{Course of action}
+ & 1. Use the steps listed above in order to start the SUT. \\
+ & 2. Execute the script by running \texttt{\# code/client/tests/6.py} \\\hline
+ Valid trace & Verify that the script prints 'Success'. \\\hline
+ \hline
+
+
+ Nr & 6 \\\hline
+ Title & Request with corrupted destination port. \\\hline
+ Input & Generated packets with 1byte payload, in these packets the destination port number is increased by one. \\\hline
+ Expected output & - \\\hline
+ \multirow{2}{*}{Course of action}
+ & 1. Use the steps listed above in order to start the SUT. \\
+ & 2. Execute the script by running \texttt{\# code/client/tests/6.py} \\\hline
+ Valid trace & Verify that the script prints 'Success'. \\\hline
+ \hline
+
+\end{longtable}
+
+%\begin{tabularx}{\linewidth}{| l | X|}
+%\hline
+%Nr & 1 \\\hline
+%Title & Single valid request. \\\hline
+%Input & Pcap file with prerecorded valid packets. \\\hline
+%Expected output & Pcap file with valid response to request. \\\hline
+%Course of action & \begin{enumerate}
+% \item Execute \emph{./scripts/tests/case1-single-valid.sh}
+% \item Load \emph{output/case1.pcap} with ...
+%\end{enumerate} \\\hline
+%Valid trace & \begin{enumerate}
+% \item \textbf{Hier packets benoemen?}
+%\end{enumerate} \\\hline
+%\end{tabularx}
+%
+%\begin{tabularx}{\linewidth}{| l | X|}
+% \hline
+% Nr & 2 \\\hline
+%Title & Single request with corrupted checksum. \\\hline
+% Input & Pcap file used as \emph{test-case 1} input. \\\hline
+% Expected output & No response from SUT, logs with rejected packets. \\\hline
+% Course of action & \begin{enumerate}
+% \item Load input pcap file into ....
+% \item Corrupt checksum of loaded packets.
+% \item Save resulting packets as pcap file.
+% \item Load new pcap file into ...
+% \item Replay new pcap file.
+% \item Record SUT response using...
+% \item Extract log with rejected packets.
+% \item Save recorded packets as a pcap file.
+% \item Analyze packets in resulting file.
+% \end{enumerate} \\\hline
+% Valid trace & \begin{enumerate}
+% \item \textbf{Aangeven welke packets corrupted zijn?}
+% \end{enumerate} \\\hline
+%\end{tabularx}
+%\end{table}
repositories / tt2015.git / blobdiff