Skip to content
This repository has been archived by the owner on Dec 16, 2022. It is now read-only.

Commit

Permalink
work on report, many sections
Browse files Browse the repository at this point in the history
  • Loading branch information
@taivop
taivop committed Dec 6, 2016
1 parent 8385ddf commit 0981e0e
Show file tree
Hide file tree
Showing 2 changed files with 49 additions and 40 deletions.
Binary file modified report/milestone3.pdf
View file
Binary file not shown.
89 changes: 49 additions & 40 deletions report/milestone3.tex
View file
Original file line number Diff line number Diff line change
Expand Up @@ -13,6 +13,14 @@
\pagestyle{fancy}
\fancyhf{}

\fancypagestyle{firstpagefooter}
{
\lfoot{Compiled: \today}
\cfoot{}
\rfoot{\thepage}
}
\cfoot{\thepage}


\renewcommand{\headrulewidth}{0pt}
\renewcommand{\footrulewidth}{0pt}
Expand Down Expand Up @@ -50,36 +58,26 @@
}

\maketitle
\thispagestyle{firstpagefooter}
\newpage

\tableofcontents


\section*{Notes on writing the report}

The report does not need to be extensive but it must be concise, complete, and correct. Conciseness is important in terms of content and explanations, focusing on what has been done and explanations of the results. A long report is not necessarily a better report, especially if there are aspects that remain unexplained. Completeness implies that the report should give a comprehensive idea of what has been done by mentioning all key aspects of the modeling and analysis effort. Limited analysis because of flaws in the system or lack of experimental data from Milestones 1 or 2 are not valid arguments for an incomplete report. If bugs or lack of data prevent you from doing a correct analysis, the system must be debugged and new data collected. In case the system has been modified, include a short description of the changes as an appendix.

Remember that this is a report about modeling and analyzing the system you have designed and built, using the experimental data you have collected. There is no unique way to do the report and you may choose to focus on different aspects of the system as long as you deliver a complete analysis of its behavior. Keep in mind that, \emph{for all queuing models in the report}, you need to explain how the the parameters of the model were determined and from which experiments the data comes from (adding a reference to the exact graph, table, etc. from the previous milestones). You have to find all system metrics that can be derived using the corresponding formulas and then match to the experimental results, explaining the similarities and differences in quantitative and qualitative terms. The calculations and the numbers you derive might need to be explained with references to the logs and sources in the previous reports. Make sure to mark these references, as well as the ones pointing to experimental results clearly. \textit{Missing parts of the above requirements might lead to significant loss of points in each section.}

The report should be organized in sections as explained in the next pages, and each section should address at least the questions mentioned for each point. You might be called for a meeting in person to clarify aspects of the report or the system and to make a short presentation of the work done. By submitting the report, you confirm that you have done the work on your own, the data used comes from experiments your have done, you have written the report on your own, and you have not copied neither text nor data from other sources.

\medskip
The milestone is worth 200 points.

\pagebreak

\clearpage
% --------------------------------------------------------------------------------
% --------------------------------------------------------------------------------
\section{System as One Unit}\label{sec:system-one-unit}
\section{System as One Unit}\label{sec:part1-system-one-unit}
% --------------------------------------------------------------------------------
% --------------------------------------------------------------------------------

\subsection{Data}

The experimental data used in this section comes from the updated trace experiment, found in \texttt{\href{https://gitlab.inf.ethz.ch/pungast/asl-fall16-project/tree/master/results/trace\_rep3}{results/trace\_rep3}}. For details, see Milestone 2, Appendix A. The first 2 minutes and last 2 minutes were dropped as warm-up and cool-down time similarly to previous milestones.
The experimental data used in this section comes from the updated trace experiment, found in \texttt{\href{https://gitlab.inf.ethz.ch/pungast/asl-fall16-project/tree/master/results/trace\_rep3}{results/trace\_rep3}} (short names \texttt{trace\_ms*}, \texttt{trace\_mw} and \texttt{trace\_req} in Milestone~1). For details, see Milestone~2, Appendix A.

The first 2 minutes and last 2 minutes were dropped as warm-up and cool-down time similarly to previous milestones.

\subsection{Model}
\label{sec:part1:model}

In this section I create an M/M/1 model of the system. This means the following definitions and assumptions:
\begin{itemize}
Expand All @@ -99,6 +97,8 @@ \subsection{Model}

\subsection{Comparison of model and experiments}

\todo{this whole subsection}

\input{../results/analysis/part1_mm1/comparison_table.txt}

\begin{figure}[h]
Expand All @@ -117,32 +117,45 @@ \subsection{Comparison of model and experiments}
\clearpage
% --------------------------------------------------------------------------------
% --------------------------------------------------------------------------------
\section{Analysis of System Based on Scalability Data}\label{sec:analysis-scalability}
\section{Analysis of System Based on Scalability Data}\label{sec:part2-analysis-scalability}
% --------------------------------------------------------------------------------
% --------------------------------------------------------------------------------

\subsection{Guidelines}
Length: 1-4 pages
\subsection{Data}

The experimental data used in this section comes from \todo{Section~x} and can be found in \texttt{\href{https://gitlab.inf.ethz.ch/pungast/asl-fall16-project/tree/master/results/trace\_rep3}{\todo{}}}.

Starting from the different configurations that you used in the second milestone, build M/M/m queuing models of the system as a whole. Detail the characteristics of these series of models and compare them with experimental data. The goal is the analysis of the model and the real scalability of the system (explain the similarities, the differences, and map them to aspects of the design or the experiments). Make sure to follow the model-related guidelines described in the Notes!

\subsection{Model}

\todo{} describe the model I was using, and how I found the parameters
The assumptions and definitions of the M/M/m model are the same as for the M/M/1 model laid out in Section~\ref{sec:part1:model} with the following modifications:

\begin{itemize}
\item We treat the SUT as a collection of $m$ servers.
\item All jobs waiting for service are held in one queue.
\item If any server is idle, an arriving job is serviced immediately.
\item If all servers are busy, an arriving job is added to the queue.
\end{itemize}

\paragraph{Parameters}
\todo{} describe how I found the parameters

bad model because I map requests to servers uniformly. M/M/m assumes that each server takes a request when it finishes with the previous one, but that is not true in my case.

\subsection{Data}
\paragraph{Problems}
\todo{}

The experimental data used in this section comes from \todo{Section x} and can be found in \texttt{\href{https://gitlab.inf.ethz.ch/pungast/asl-fall16-project/tree/master/results/trace\_rep3}{\todo{}}}.
\begin{enumerate}
\item I actually have $m$ queues (one for each server), not a single queue; each request is assigned to a server when \linkmain{LoadBalancer} receives it.
\item I map requests to servers uniformly. M/M/m assumes that each server takes a request when it finishes with the previous one, but that is not true in my case -- I take earlier
\end{enumerate}

\subsection{Comparison of model and experiments}


\clearpage
% --------------------------------------------------------------------------------
% --------------------------------------------------------------------------------
\section{System as Network of Queues}\label{sec:network-of-queues}
\section{System as Network of Queues}\label{sec:part3-network-of-queues}
% --------------------------------------------------------------------------------
% --------------------------------------------------------------------------------

Expand All @@ -155,7 +168,7 @@ \subsection{Guidelines}
\clearpage
% --------------------------------------------------------------------------------
% --------------------------------------------------------------------------------
\section{Factorial Experiment}\label{sec:2k-experiment}
\section{Factorial Experiment}\label{sec:part4-2k-experiment}
% --------------------------------------------------------------------------------
% --------------------------------------------------------------------------------

Expand All @@ -167,41 +180,37 @@ \subsection{Guidelines}
\clearpage
% --------------------------------------------------------------------------------
% --------------------------------------------------------------------------------
\section{Interactive Law Verification}\label{sec:interactive-law}
\section{Interactive Law Verification}\label{sec:part5-interactive-law}
% --------------------------------------------------------------------------------
% --------------------------------------------------------------------------------

\subsection{Guidelines}
Length: 1-2 pages

Check the validity of all experiments from one of the three sections in your Milestone 2 report using the interactive law (choose a section in which your system has at least 9 different configurations). Analyze the results and explain them in detail.

\subsection{Data}

The experimental data used in this section comes from Milestone 2, Section 2 (Effect of Replication) and can be found in \texttt{\href{https://gitlab.inf.ethz.ch/pungast/asl-fall16-project/tree/master/results/replication}{results/replication}}.
The experimental data used in this section comes from Milestone~2, Section~2 (Effect of Replication) and can be found in \texttt{\href{https://gitlab.inf.ethz.ch/pungast/asl-fall16-project/tree/master/results/replication}{results/replication}} (short name \texttt{replication-S*-R*-r*} in Milestone~2).

\subsection{Results}
The first 2 minutes and last 2 minutes were \textbf{not} dropped because IRTL should hold also in warm-up and cool-down periods. Repetitions at the same configuration were considered as separate experiments.

which experiment I chose, did I remove beginning and end, ...
\subsection{Results}

mention mean error
mention value of mean error (mean of Figure~\ref{fig:part5:error_percentage})

\todo{also show graphs as function of parameters (like in ms2 exp2)?}

\todo{why do I predict lower throughput?} because total cycle time is higher than it should for given throughput -- but why? since measuring is done by memaslap, probably the problem lies on that side

\begin{figure}[h]
\centering
\begin{subfigure}[b]{0.49\textwidth}
\begin{subfigure}[t]{0.49\textwidth}
\includegraphics[width=\textwidth]{../results/analysis/part5_irtl/graphs/predicted_vs_actual_throughput.pdf}
\caption{ \todo{} }
\caption{Throughput predicted using IRTL, as a function of actual throughput calculated from experimental data. Note the horizontal scale does not include 0.}
\label{fig:part5:predicted_vs_actual}
\end{subfigure}
\begin{subfigure}[b]{0.49\textwidth}
\begin{subfigure}[t]{0.49\textwidth}
\includegraphics[width=\textwidth]{../results/analysis/part5_irtl/graphs/error_percentage.pdf}
\caption{ \todo{} . note scale}
\caption{Histogram of the relative error of throughput predicted using IRTL, counting the number of experiments in a given error range. Note the horizontal scale does not include 0.}
\label{fig:part5:error_percentage}
\end{subfigure}
\caption{Evaluation of the validity of Milestone~2 Section~2 experiments}
\end{figure}


Expand Down

0 comments on commit 0981e0e

Please sign in to comment.