-
Notifications
You must be signed in to change notification settings - Fork 0
/
solution.tex
50 lines (37 loc) · 2.11 KB
/
solution.tex
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
\section{Our Approach}
\label{s:solution}
These requirements seem complex and even self-contradictory, but
computer science has an aphorism: ``We can solve any problem by
introducing an extra level of indirection''.
In our solution, metadata provides that level of indirection.
Our solution integrates three contributions to address these
requirements, with the key idea being the decoupling of publication
from data access:
\begin{enumerate}
\item An RDF vocabulary for describing log and monitoring data collections
in terms of the subject being monitored, the time period covered,
the type and format of the data and details for accessing the data or
contacting its curator. The vocabulary allows construction of a
distributed graph that can be queried to discover relevant collections
of log-like data and annotations on that data.
\item A schema for publishing annotations for log data, which can be represented in an SQL
database (or otherwise used to define the query and return fields of an API of a more complex representation) also in terms of subject and time period. Annotations
databases fit neatly into the RDF graph and might contain:
\begin{itemize}
\item Expert commentary, such as ``this log message means that down links
caused the network to be quiesced while re-calculating routing tables''
\item Human observations, such as ``during this period our engineer was
replacing some failed nodes, so the network was probably disrupted''
\item Machine-generated observations, such as message-pattern frequencies
computed by running certain logs through an analysis tool such as
Baler~\cite{Baler}.
\end{itemize}
\item A collection of tools to make the RDF vocabulary and annotation
schema accessible to users, system administrators and support staff.
The vocabulary and schema are independent of the tools, but
the tools provide an alternative to learning the underlying
technologies and a starting point for further tool development.
\end{enumerate}
\input{solution-01}
\input{solution-02}
\input{solution-03}