-
Notifications
You must be signed in to change notification settings - Fork 0
/
survey-work-notes.tex
39 lines (33 loc) · 1.4 KB
/
survey-work-notes.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
\subsection{Editing notes (this is not a real section and will be deleted)}
\subsubsection{Misc}
\textbf{(authors: Anders, Maxim, Mike, Peter)} \textit{Note: I mark my comments as -mxp- for brevity - Maxim}
\subsubsection{Notes on content}
\begin{itemize}
\item -mxp- there is workflow mgt and workload mgt. While they overlap at times, and also include varying degree of data management functionality, it is best to make this distinction (also based on existing literature). For that reason, I changed the title of the section.
\item -mxp- Overlap with Data Management.
\item -mxp- DIRAC, PanDA, CRAB3/ Phedex.
\end{itemize}
\subsubsection{Boilerplate by Brett (to be removed)}
Possible sub-categories
\begin{description}
\item[Processing] issues:
\begin{itemize}
\item Submission tools:
\item Running multiple instances of a job differing only by parameters or input files -mxp- folds into workflow mgt
\item Running jobs in parallel on local batch or wider Grid
\item Workflow engines
\end{itemize}
\item[Monitoring] issues:
\begin{itemize}
\item Overall health, throughput, usage, abuse
\item Locating and understanding failures
\item Determining systemic problems
\item Drilling down to details at the level of individual jobs
\end{itemize}
\item[Recovery] issues:
\begin{itemize}
\item rerun failed jobs
\item clean up droppings
\item fix systemic problems
\end{itemize}
\end{description}