diff options
Diffstat (limited to 'main.tex')
| -rw-r--r-- | main.tex | 129 |
1 files changed, 107 insertions, 22 deletions
@@ -2,11 +2,11 @@ \input{style/style.tex} \begin{document} -\frame{\titlepage \centering \footnotesize Online slides: \url{https://www.overleaf.com/read/tknhxmqfgtdy\#87413e}} +\frame{\titlepage \centering \footnotesize Online slideshow: \url{mjkw.pl/vu/bsc}} \begin{frame}\frametitle{Motivation} \begin{tcolorbox}[title=Context] - 21\textsuperscript{st} century datacenters are primarily heterogeneous~\cite{DBLP:conf/date/MilojicicFDR21} and + 21\textsuperscript{st} century datacenters (DC) are mostly heterogeneous~\cite{DBLP:conf/date/MilojicicFDR21} and modern computational needs of AI drive managers to diversify datacenters even more~\cite{DBLP:journals/computer/AthavaleBBMMPS24}. In result datacenters become extremely complex and hard to operate with millions of CPU's, GPU's etc. \end{tcolorbox} @@ -19,12 +19,11 @@ Left to right: a Google datacenter, server racks, Ada Lovelace AD102 GPU architecture. \end{frame} - \begin{frame}\frametitle{Problem Statement} \begin{tcolorbox}[title=DCDT's lack predictive analytics] We need Datacenter Digital Twins (DCDT) to be better able to detect and solve issues in critical ICT infrastructure~\cite{DBLP:journals/computer/AthavaleBBMMPS24}. However, DCDT's are still actively developed and lack crucial features such as predictive analytics~\cite{DBLP:usdoe/report/AP26894} to \emph{e.g.,} prevent unexpected failures. - With predictive analysis (\emph{e.g.,} regression) DCDT's could save millions of lost \$USD~\cite{DBLP:conf/acsos/TalluriOVTI21}. + With predictive analysis (\emph{e.g.,} simulation) DCDT's could save millions of lost \$USD~\cite{DBLP:conf/acsos/TalluriOVTI21}. \end{tcolorbox} \begin{center} @@ -33,7 +32,7 @@ \tiny \textbf{Figure 1.2:} Where does our work fit within the field of datacenter digital twinning? There are 5 core elements to any Digital Twin: \myCircled{A} The Digital $\rightarrow$ Physical Twin link, \myCircled{B} the Physical Twin (\emph{e.g.,} the datacenter), \myCircled{C} the Physical $\rightarrow$ Digital Twin link, \myCircled{D} the Digital Twin, \myCircled{E} the features necessary to any Digital Twin. - \textcolor{ForestGreen}{\faHighlighter~Highlighted areas are the contributions from this thesis, which include the autonomous actions resulting from predictive insights \myCircledGreen{A} and the predictive analysis itself within \myCircledGreen{E}.} + \textcolor{Green}{\faHighlighter~Highlighted areas are the contributions from this thesis, which include the autonomous actions resulting from predictive insights \myCircledGreen{A} and the predictive analysis itself within \myCircledGreen{E}.} \end{frame} \begin{frame}\frametitle{Research Questions} @@ -46,7 +45,7 @@ \end{tcolorbox} \begin{tcolorbox}[title=Research Question 2] - How to design a datacenter digital twin reference architecture using discrete-event simulation and predictive data analytics? + How to design a reference architecture for a predictive datacenter digital twin using discrete-event simulation? \end{tcolorbox} \begin{tcolorbox}[title=Research Question 3] @@ -57,45 +56,131 @@ \begin{frame}\frametitle{\textbf{RQ1}: Literature Review I} \begin{tcolorbox}[title=Results] - This is a dummy sentence meant to make the tcolorbox have more than 2 lines of text width so that I am able to show the text and the table spacing better. - I hope it fits its purpose well. + The literature on DCDTs is scarce. + Some systems barely classify as DTs (\emph{e.g.,} Kalibre~\cite{DBLP:conf/sensys/WangZD0TCWZ20}, ChatTwin~\cite{DBLP:conf/sensys/LiW0Z0T23}). + Existing deployments specialize in \textcolor{Red}{Cooling and Heat Modelling}, together with \textcolor{Red}{3D visualizations}. + Most lack crucial predictive DC behaviour modelling. \end{tcolorbox} \input{images/table.tex} + % Research on DTs for datacenters have been separate, siloed efforts focused on either datacenter cooling, network performance, power consumption or visualization efforts. + % CFD usually means Navier-Stokes equations. + % CFD models take ages to compute. \end{frame} \begin{frame}\frametitle{\textbf{RQ1}: Literature Review II} % Mandatory: split the figure into 2: top and bottom, and that way you can fill in the entire slide nicely. + + \begin{tcolorbox}[title=A holistic DCDT system model] + We propose a generic model of datacenter digital twinning that can be mapped to each system from \textbf{Table 1.1}. To answer \textbf{RQ2}, we design a ref. arch. for \emph{Operations Model}. + We introduce the \emph{Digital Thread}: a bridge between software and reality. + \end{tcolorbox} + \begin{center} + \vspace{-0.1cm} + \includegraphics[width=0.8\textwidth]{images/system_model2.pdf} + \end{center} + % The reason why the cooling system is in the graph is because of the fact that 40\% of total energy consumed in DCs comes from cooling~\cite{DBLP:conf/noms/ZhangZLZWC22}. + % It has come to the point where datacenters are being build in the Pan-Arctic region, such as Finland,Russia,Sweden etc. with Iceland leading in number of DCs https://www.datacentermap.com/iceland/ + % The SmarDC digital twin is purely to get more training data for AI models. + % Not really a digital twin per se. + + \tiny + \textbf{Figure 1.3:} To answer \textbf{RQ1} we designed a generic datacenter digital twin system model based on a comprehensive literature review and findings from \textbf{Table 1.1}. The \emph{Infrastructure Model} simulates the structure of the DC and the \emph{Operations model} simulates the behaviour of the DC. + % Consider splitting the figure into 2 a.k.a. top and bottom. + % By the AIAA definition, the DT mimicks the structure and behaviour. + % Data Lake -> Data Storage + % Use cases of DT's found by Brewer et al.: augmented reality, forensic analysis and diagnostics, predictive modelling, failure detection, operational optimization, ``what-if''' scenarios and virtual prototyping. +\end{frame} + +\begin{frame}\frametitle{\textbf{RQ2}: Reference Architecture} + % Make Kafka logos clearly defined --> add a legend with icons? + \hspace{-0.3cm} \begin{minipage}[b]{0.45\linewidth} \begin{center} - \includegraphics[width=1.15\textwidth]{images/system_model.pdf} + % Change to Datacenter (Physical Twin) + \includegraphics[width=1.15\textwidth]{images/ref_architecture.pdf} \end{center} + \vspace{-0.2cm} \tiny - \textbf{Figure 1.3:} To answer \textbf{RQ1} we designed a generic datacenter digital twin system model based on a comprehensive literature review and findings from \textbf{Table 1.1}. + \textbf{Figure 1.4:} The predictive datacenter digital twin reference architecture. + The architecture was designed with the \emph{AtLarge Design Process}~\cite{DBLP:conf/icdcs/IosupVTETBFMT19}. + \vspace{0.2cm} + \end{minipage} + \hspace{0.8cm} + \begin{minipage}[b]{0.45\linewidth} + \begin{center} + \includegraphics[width=1.15\linewidth]{images/implementation.png} + \end{center} + \vspace{-0.2cm} + \tiny + \textbf{Figure 1.5:} The prototype based on \textbf{Figure 1.4} towards answering \textbf{RQ3}. + The time-series data flows first to the \texttt{Grafana} dashboard, \texttt{PostgreSQL} database and \texttt{Redis} cache, as advised in~\cite{DBLP:conf/sc/TaheriBPRHDEWPM24}. \end{minipage} - % Consider splitting the figure into 2 a.k.a. top and bottom. - % Data Lake -> Data Storage -\end{frame} -\begin{frame}\frametitle{\textbf{RQ2}: Reference Architecture} + % We decided to use discrete-event simulation, as opposed to computational fluid dynamics because of the high overheads of development time needed for CFD. + % CFD simply takes too long to run, making it unfeasible for real-time analytics and simulation. + % Citing ExaDigit: [CFD] they are also more computationally expensive, generally making real-time operation unfeasible. + % Consider adding this minipage directly to the ``draw.io'' diagram +\end{frame} +% You should skip \hfill completely or in favour of \hspace very minimally. +\begin{frame}\frametitle{\textbf{RQ3}: Experimental Setup} + \begin{minipage}[b]{0.45\linewidth} + \begin{tcolorbox}[title=Problem, colbacktitle=red!70!black,colback=red!20!white] + We cannot just go and test digital twins on large systems, because we do not have large systems at hand. + Moreover, real-world experimentation is costly and unsustainable in the long run~\cite{DBLP:conf/ccgrid/MastenbroekAJLB21}. + \end{tcolorbox} + \vspace{0.5cm} + \begin{tcolorbox}[title=Solution, colbacktitle=Green!70!black, colback=Green!20!white] + \scriptsize + They way we test our reference architecture prototype is by using multiple simulators. + We use an additional OpenDC process to play the role of a real datacenter. + \end{tcolorbox} + \vspace{1cm} + \end{minipage} + \hspace{0.25cm} + \begin{minipage}[b]{0.45\linewidth} + \vspace{-0.2cm} + \begin{center} + \includegraphics[width=1.2\linewidth]{images/predictive_analyticsv3.pdf} + \end{center} + \tiny + \vspace{-0.2cm} + \textbf{Figure 1.6:} The experimental setup. + Answering \textbf{RQ3} we provide a novel way to evaluate datacenter digital twins through discrete-event simulation. + \end{minipage} \end{frame} \begin{frame}\frametitle{\textbf{RQ3}: Experimental Results I} - \begin{tcolorbox}[title=Main Finding I] - Here explain what did you find. + % You have some model, and this can be based on multiple traces. + %Get insight from CINECA --> you get a probability of certain hosts failing. + % Anomaly detection --> CINECA, how good their detection is? + %If you incorporate that? If you can make the case that because of our new digital twin we can incorporate such models, anomaly/failure detection, from CINECA. + %If we had that in, we can reach these kinds of gains. + % @Mateusz there is really not a possibility to incorporate CINECA's models, so to address Dante's feedback, I created this experiment. + + \begin{tcolorbox}[title=Failure Detection: Main Finding I] + On average, \emph{Sunfish} can detect 14.5\% of unexpected failures in the physical twin. + We show, that digital twinning \emph{can} be used for failure detection. + \end{tcolorbox} - Here goes the figure that backs up claim in Main Finding I. - Evidence for Main Finding I. + \begin{minipage}[b]{0.45\linewidth} + \begin{center} + \includegraphics[width=1.1\textwidth]{images/23_Jun_2026_102028.pdf} + \end{center} + \vspace{-0.3cm} + \tiny + \textbf{Figure 1.5:} Experiment 1 Setup: The Digital Twin estimates the failures based on the Normal Distribution \emph{N\textasciitilde($\mu$,$\sigma$)} with $\mu = 1.5$ and $\sigma = 0.5$. + ``Real'' OpenDC failures come from a WhatsApp user reports. + \end{minipage} % Explain what the axis are in the figure caption. % Talk about the experimental setup in the figure. % Give more reliable results than just numbers -- do statistical testing, i.e., standard deviation, confidence intervals. \end{frame} - \begin{frame}\frametitle{\textbf{RQ3}: Experimental Results II} - \begin{tcolorbox}[title=Main Finding II] + \begin{tcolorbox}[title=Scheduling Optimization: Main Finding II] Here explain what did you find. \end{tcolorbox} - Here goes the figure that backs up claim in Main Finding II. + \end{frame} \begin{frame}\frametitle{Key Takeaways} @@ -121,7 +206,7 @@ \end{frame} -\setcounter{framenumber}{3} +\setcounter{framenumber}{4} \setbeamertemplate{footline}[page number]{ % Unfortunately this must remain here. |