Importing 662 stuff
parent
6269ed6b05
commit
f4b7c262dd
Binary file not shown.
Binary file not shown.
Binary file not shown.
Binary file not shown.
After Width: | Height: | Size: 68 KiB |
Binary file not shown.
Binary file not shown.
Binary file not shown.
Binary file not shown.
After Width: | Height: | Size: 59 KiB |
11
main.bib
11
main.bib
|
@ -1,13 +1,22 @@
|
|||
%% This BibTeX bibliography file was created using BibDesk.
|
||||
%% http://bibdesk.sourceforge.net/
|
||||
|
||||
%% Created for Oliver Kennedy at 2016-01-17 23:38:39 -0500
|
||||
%% Created for Oliver Kennedy at 2016-01-18 15:23:25 -0500
|
||||
|
||||
|
||||
%% Saved with string encoding Unicode (UTF-8)
|
||||
|
||||
|
||||
|
||||
@misc{ramamurthy2015pocketdata,
|
||||
Author = {Naveen Kumar Ramamurthy and Sankara Vadivel Dhandapani and Saravanan Adaikkalavan and Sathish Kumar Deivasigamani},
|
||||
Date-Added = {2016-01-18 20:07:55 +0000},
|
||||
Date-Modified = {2016-01-18 20:10:15 +0000},
|
||||
Howpublished = {University at Buffalo CSE-662 Final Report},
|
||||
Month = {Dec},
|
||||
Title = {PocketData Benchmark (Team TBD)},
|
||||
Year = {2015}}
|
||||
|
||||
@article{Klonatos:2014:BEQ:2732951.2732959,
|
||||
Acmid = {2732959},
|
||||
Author = {Klonatos, Yannis and Koch, Christoph and Rompf, Tiark and Chafi, Hassan},
|
||||
|
|
|
@ -58,25 +58,25 @@ We noted above that a large proportion of \texttt{SELECT} queries were exact loo
|
|||
\begin{subfigure}[t]{0.33\textwidth}
|
||||
\includegraphics[width=\textwidth]{./graphs/activity/All_Devices__Top_10__All_Queries__ByAppPreviousQueryCDFGraph_noKey.pdf}
|
||||
\caption{}
|
||||
\label{fig-app-interarrival}
|
||||
\label{fig:app:interarrival}
|
||||
\end{subfigure}%
|
||||
\begin{subfigure}[t]{0.33\textwidth}
|
||||
\includegraphics[width=\textwidth]{./graphs/activity/All_Devices__Top_10__All_Queries__ByAppRuntimeCDFGraph_noKey.pdf}
|
||||
\caption{}
|
||||
\label{fig-app-runtime}
|
||||
\label{fig:app:runtime}
|
||||
\end{subfigure}%
|
||||
\begin{subfigure}[t]{0.33\textwidth}
|
||||
\includegraphics[width=\textwidth]{./graphs/activity/All_Devices__Top_10__All_Queries__ByAppRowcountCDFGraph_noKey.pdf}
|
||||
\caption{}
|
||||
\label{fig-app-rowcount}
|
||||
\label{fig:app:rowcount}
|
||||
\end{subfigure}%
|
||||
\caption{Per-App Summary Statistics for Android SQLite Queries. Distributions of (a) inter-query arrival times, (b) query runtimes, and (c) rows returned per query.}
|
||||
\label{fig-app}
|
||||
\label{fig:app}
|
||||
\end{figure*}
|
||||
|
||||
Figure~\ref{fig-app} shows query interarrival times, runtimes, and returned row
|
||||
Figure~\ref{fig:app} shows query interarrival times, runtimes, and returned row
|
||||
counts for ten of the most active SQLite clients. As seen in
|
||||
Figure~\ref{fig-app-interarrival}, the 0.01Hz periodicity is not unique to any one
|
||||
Figure~\ref{fig:app:interarrival}, the 0.01Hz periodicity is not unique to any one
|
||||
application, further suggesting filesystem locking as a culprit. Two of the most
|
||||
prolific SQLite clients, \textit{Google Play services} and \textit{Media Storage}
|
||||
appear to be very bursty: 70\% of all statements for these applications are issued
|
||||
|
@ -85,13 +85,13 @@ issued by the \textit{Android System} itself. The interarrival time CDF appears
|
|||
to be almost precisely logarithmic for rates above 10$\mu$s, but has a notable lack
|
||||
of interarrival times in the 1ms to 10ms range. This could suggest caching
|
||||
effects, with the cache expiring after 1ms.
|
||||
As seen in Figure~\ref{fig-app-runtime}, most apps hold to the average runtime of
|
||||
As seen in Figure~\ref{fig:app:runtime}, most apps hold to the average runtime of
|
||||
100$\mu$s, with several notable exceptions. Over 50\% of the
|
||||
\textit{Android System}'s statements take on the order of 1ms. Just under 20\% of
|
||||
\textit{Hangouts} statements take 10ms, suggesting an update-heavy workload. Also,
|
||||
\textit{Contacts Storage} has a heavier-duty workload, with 30\% of statements taking
|
||||
between 100$\mu$s and 1ms.
|
||||
Figure~\ref{fig-app-rowcount} shows that the \textit{Android System} and
|
||||
Figure~\ref{fig:app:rowcount} shows that the \textit{Android System} and
|
||||
\textit{Media Storage} issue almost exclusively single-row lookup queries.
|
||||
The remaining apps issue a large number of single-row queries --- Even
|
||||
\textit{Contacts Storage} has a workload consisting of 45\% single-row reads ---
|
||||
|
@ -101,6 +101,25 @@ at a time, suggesting that these views are backed directly by SQLite. Although
|
|||
apps have long tails, two apps in particular: \textit{Gmail} and \textit{Google+} are
|
||||
notable for regularly issuing queries that return on the order of 100 rows.
|
||||
|
||||
\begin{figure*}
|
||||
\centering
|
||||
\begin{subfigure}[t]{0.5\textwidth}
|
||||
\includegraphics[width=\textwidth]{graphs/whatsapp-minified.pdf}
|
||||
\caption{}
|
||||
\end{subfigure}%
|
||||
\begin{subfigure}[t]{0.5\textwidth}
|
||||
\includegraphics[width=\textwidth]{graphs/facebook-minified.pdf}
|
||||
\caption{}
|
||||
\end{subfigure}%
|
||||
\caption{Variations in bursty data access patterns~\cite{ramamurthy2015pocketdata} for WatsApp (a) and Facebook (b).}
|
||||
\label{fig:burstiness}
|
||||
\end{figure*}
|
||||
|
||||
Figure~\ref{fig:burstiness} shows variations in query burstiness across multiple apps and users\footnote{The PIs have already incorporated material from this proposal into their coursework. Figure \ref{fig:burstiness} is from a student report~\cite{ramamurthy2015pocketdata} from UB's CSE-662, jointly instructed by PIs Kennedy and Ziarek.}.
|
||||
Two features immediately emerge from this data.
|
||||
First, \PocketData{} workloads are extremely bursty; The default steady state is completely idle, with infrequent bursts of hundreds of operations per second.
|
||||
Second, the nature of these bursts varies significantly by the calling app; In this trace Facebook generates a read-only workload, while Watsapp produces two bursts each with a distinct mix of updates, inserts, deletes, and selects.
|
||||
|
||||
\medskip
|
||||
|
||||
We plan to freely releasing aggregate metrics about database usage patterns in
|
||||
|
|
|
@ -99,10 +99,10 @@ systems and query processing systems, in embedded deployments. The programming l
|
|||
community is exploring domain specific languages for specialized query processing.
|
||||
The mobile community is continually exploring how to push the envelope on smartphone based computing,
|
||||
whether via power aware mechanisms, or through more adaptive systems. Many of these solutions use mobile databases
|
||||
as their fundamental computation engine (e.g. the {\em Maybe} system developed at UB) or must
|
||||
as their fundamental computation engine (e.g. the `\texttt{maybe}' system developed at UB) or must
|
||||
consider the performance characteristics of database systems (e.g. power modeling).
|
||||
|
||||
This proposal aims to create a community research around our \PocketData{} toolchain
|
||||
This proposal aims to create a community research infrastructure around our \PocketData{} toolchain
|
||||
to enable a myriad of research activities for above mentioned communities. Additionally,
|
||||
in this planning grant, we will explore the precise needs of these communities to ensure an
|
||||
infrastructure that has broad applicability.
|
||||
|
|
Loading…
Reference in New Issue