diff options
| -rw-r--r-- | content/conclusion.tex | 23 | ||||
| -rw-r--r-- | sources/dt_features_comparison.tex | 20 |
2 files changed, 21 insertions, 22 deletions
diff --git a/content/conclusion.tex b/content/conclusion.tex index f560bc6..122b2be 100644 --- a/content/conclusion.tex +++ b/content/conclusion.tex @@ -2,24 +2,23 @@ Datacenter manageability is a top-priority for the digital society. Over 3 million jobs in the Netherlands directly depend on cloud services, which are hosted in datacenters~\cite{DBLP:journals/corr/IosupKLVG22}. -Datacenter digital twinning, a promising management technique can offer unique insight into complex facility behaviour. -In this thesis we pave the way to more advanced \gls{dcdt}s. +Datacenter digital twinning, a promising management technique can offer unique insight into complex facility behaviour~\cite{DBLP:journals/computer/AthavaleBBMMPS24}. +In this thesis we paved the way to more advanced \gls{dcdt}s. We contribute to the scientific community a set of findings that we hope will prove helpful in enabling predictive analytics in both existing \gls{dcdt}s and future projects. Starting from a thorough investigation into the new, emerging field of datacenter digital twinning, we designed a system capable of incorporating sophisticated data analysis techniques. We ended our project with a novel evaluation method used in a set of exhaustive experiments. -As such, we believe we answer the main research question by addressing each sub-research question. +We answer the main research question by addressing each sub-research question. \section{Answers to Research Questions}\label{ss:answers_to_rqs} \begin{enumerate}[label=\emph{RQ\textsubscript{\arabic*}}] \item \emph{How to asses the current state-of-the-art of digital twinning for datacenters?}\\ In order to answer this research question, we conducted a semi-structured literature review. - Our findings indicate that the field of datacenter digital twinning is still under development. - There exist few existing \gls{dcdt} deployments. - The current efforts in modelling datacenters focus on very specialized parts of the datacenter management, \ie cooling and heat modelling, network mapping. - These standalone systems fail to offer the holistic capabilities envisioned for \gls{dt}s. - The results of the literature survey are in \Cref{tab:dt_features_comparison}. - \Cref{tab:dt_features_comparison} contains systems which we found though a semi-structured literature review process. + Our findings indicate that the field of datacenter digital twinning is still under development, and there exist few \gls{dcdt} deployments. + The current efforts in modelling datacenters focus on very specialized parts of datacenter management, \ie cooling and heat modelling, network mapping. + Many crucial features, inherent to the \gls{dt} definition are still missing from current \gls{dcdt}s. + Present, standalone \gls{dcdt} systems fail to offer the holistic capabilities envisioned by the inventors of \gls{dt}s. + The results of the literature survey are in \Cref{tab:dt_features_comparison}, which contains systems which we found through a semi-structured literature review process. We first used structured queries, followed by a mix of snowballing and manual search. As a result, the second contribution to answering research question 2 is a holistic system model that encompasses the features of all the systems from \Cref{tab:dt_features_comparison} (see \Cref{fig:system_model}). @@ -33,13 +32,13 @@ As such, we believe we answer the main research question by addressing each sub- \item \emph{How to validate and evaluate a datacenter digital twin architecture in relation to system requirements?}\\ To answer the last research question we crated a prototype to evaluate our system. Lacking the physical datacenter to experiment with, we came up with a novel digital twin evaluation method, that uses discrete-event simulation to model the physical datacenter. - Our main findings indicate that \gls{my_system} can reliably differentiate between large host failures and insignificant single host downtime using predictions based on the results from \code{OpenDC}, a state of the art datacenter modelling software. - Moreover, we show that \gls{my_system} can be used to incorporate a predictive analytics system and significantly lower the total number of task failures during a workload. + Our main findings indicate that \gls{my_system} can reliably differentiate between large host failures and insignificant downtime using predictions based on the results from \code{OpenDC}, a state of the art datacenter modelling software. + Moreover, we show that \gls{my_system} can be used to incorporate predictive analytics systems and significantly lower the total number of task failures during a workload. \end{enumerate} \section{Future Work}\label{ss:future_work} -We envision \gls{dcdt}s as systems that encompass features necessary to model the entire datacenter behaviour. +We envision \gls{dcdt}s as systems that encompass features necessary to model the entire datacenter. It came to our attention that with the explosive growth of \gls{ai} and the diversification of datacenters under way, \gls{dt}s will be indispensable in datacenter management. To power the predictions we envision an \gls{ml}-based inference engine as a necessary component of digital twinning. The need for \gls{ml} arises naturally in scenarios where large volumes of data, requiring little to no preprocessing meet the demand for estimating future facility behaviour. diff --git a/sources/dt_features_comparison.tex b/sources/dt_features_comparison.tex index 143ae3e..788382d 100644 --- a/sources/dt_features_comparison.tex +++ b/sources/dt_features_comparison.tex @@ -2,26 +2,26 @@ \resizebox{\columnwidth}{!}{ \begin{tabular}{lm{0.2\linewidth}m{0.25\linewidth}m{0.3\linewidth}l} \toprule - Project & Simulation Technique & Focus & Stakeholders & Modelling Capability \\ + Project & Simulation Technique & Focus & Stakeholders & Modelling Capability \\ \midrule - ExaDigiT~\cite{DBLP:conf/sc/BrewerMKWBHSGGW24} & CFD/HT, AI/ML & Energy Loss\newline Prediction,\newline \fcolorbox{Red}{white}{Heat Modelling} & HPC Engineers and Operators & \fcolorbox{Red}{white}{3D\textsuperscript{$\star$}}, \fcolorbox{Red}{white}{CH\textsuperscript{$\ddagger$}}, VP\textsuperscript{$\star$}, PE\textsuperscript{$\ddagger$}, RA, SE\textsuperscript{$\dagger$} \\ + ExaDigiT~\cite{DBLP:conf/sc/BrewerMKWBHSGGW24} & CFD/HT, AI/ML & Energy Loss\newline Prediction,\newline {Heat Modelling} & HPC Engineers and Operators & {3D\textsuperscript{$\star$}}, {CH\textsuperscript{$\ddagger$}}, VP\textsuperscript{$\star$}, PE\textsuperscript{$\ddagger$}, RA, SE\textsuperscript{$\dagger$} \\ \midrule - SmartDC~\cite{DBLP:conf/noms/ZhangZLZWC22} & CFD/HT, BIM, AI/ML & \fcolorbox{Red}{white}{Heat Modelling},\newline PUE optimization & Cloud Datacenter Engineers & \fcolorbox{Red}{white}{CH\textsuperscript{$\dagger$}}, PE, \fcolorbox{Red}{white}{3D\textsuperscript{$\star$}} \\ + SmartDC~\cite{DBLP:conf/noms/ZhangZLZWC22} & CFD/HT, BIM, AI/ML & {Heat Modelling},\newline PUE optimization & Cloud Datacenter Engineers & {CH\textsuperscript{$\dagger$}}, PE, {3D\textsuperscript{$\star$}} \\ \midrule - DyTwin~\cite{DBLP:conf/sc/TaheriBPRHDEWPM24} & Gaussian Process Regression, AI/ML & Anomaly Detection & Cloud Datacenter Operators & A\textsuperscript{$\star$}, FD, VP\textsuperscript{$\star$}, SE\textsuperscript{$\dagger$} \\ + DyTwin~\cite{DBLP:conf/sc/TaheriBPRHDEWPM24} & Gaussian Process Regression, AI/ML & Anomaly Detection & Cloud Datacenter Operators & A\textsuperscript{$\star$}, FD, VP\textsuperscript{$\star$}, SE\textsuperscript{$\dagger$} \\ \midrule - ChatTwin~\cite{DBLP:conf/sensys/LiW0Z0T23} & \textbf{\footnotesize?} & Digital Twin\newline Definition Language & Cloud Datacenter Engineers & \fcolorbox{Red}{white}{ 3D\textsuperscript{$\star$}} \\ + ChatTwin~\cite{DBLP:conf/sensys/LiW0Z0T23} & \textbf{\Large \sffamily?} & Digital Twin\newline Definition Language & Cloud Datacenter Engineers & { 3D\textsuperscript{$\star$}} \\ \midrule - Reducio~\cite{DBLP:conf/sensys/CaoW0022} & POD, Gaussian\newline Process Modelling (ML) & \fcolorbox{Red}{white}{Heat Modelling} & Edge and Hyper-scale Datacenter Operators & \fcolorbox{Red}{white}{CH\textsuperscript{$\ddagger$}, 3D\textsuperscript{$\star$}}, SE \\ + Reducio~\cite{DBLP:conf/sensys/CaoW0022} & POD, Gaussian\newline Process Modelling (ML) & {Heat Modelling} & Edge and Hyper-scale Datacenter Operators & {CH\textsuperscript{$\ddagger$}, 3D\textsuperscript{$\star$}}, SE \\ \midrule - NetGraph~\cite{DBLP:conf/sigcomm/HongWDSSHZY21} & Graphs & Network Management & Cloud Datacenter Operators & VP\textsuperscript{$\star$}, RA\textsuperscript{$\star$}, - N\textsuperscript{$\star$}, SE\textsuperscript{$\dagger$} \\ + NetGraph~\cite{DBLP:conf/sigcomm/HongWDSSHZY21} & Graphs & Network Management & Cloud Datacenter Operators & VP\textsuperscript{$\star$}, RA\textsuperscript{$\star$}, + N\textsuperscript{$\star$}, SE\textsuperscript{$\dagger$} \\ \midrule - Kalibre~\cite{DBLP:conf/sensys/WangZD0TCWZ20} & CFD/HT, ML & \fcolorbox{Red}{white}{Heat Modelling} & Cloud Datacenter Engineers & \fcolorbox{Red}{white}{CH\textsuperscript{$\ddagger$}, 3D\textsuperscript{$\star$}} \\ + Kalibre~\cite{DBLP:conf/sensys/WangZD0TCWZ20} & CFD/HT, ML & {Heat Modelling} & Cloud Datacenter Engineers & {CH\textsuperscript{$\ddagger$}, 3D\textsuperscript{$\star$}} \\ \bottomrule \end{tabular} } - \caption{\textbf{Table 1.1:} Comparison of selected datacenter digital twins. \textbf{Modelling capability:} \textcolor{Red}{3D = Visualizations}; \textcolor{Red}{CH = Cooling/Heat}, PE = Power/Energy Consumption, A = Anomaly Detection, N = Network Modelling, SE = Scenario Exploration, VP = Virtual Prototyping, FD = Federation, RA = Resource Allocation; \textbf{Data Analytics}: $\ddagger$ = Predictive Analysis; $\star$ = Descriptive Analysis, $\dagger$ = Prescriptive Analysis.} + \caption{Comparison of selected datacenter digital twins. \textbf{Modelling capability:} {3D = Visualizations}; {CH = Cooling/Heat}, PE = Power/Energy Consumption, A = Anomaly Detection, N = Network Modelling, SE = Scenario Exploration, VP = Virtual Prototyping, FD = Federation, RA = Resource Allocation; \textbf{Data Analytics}: $\ddagger$ = Predictive Analysis; $\star$ = Descriptive Analysis, $\dagger$ = Prescriptive Analysis.} \label{tab:dt_features_comparison} \end{table} % Autonomous decisions (autonomous twinning) from~\cite{DBLP:conf/sc/BrewerMKWBHSGGW24} is inherent to digital twinning -- it is better left unsaid in the table. |