Merge pull request #7 from lsst/tickets/PREOPS-576

new template, minor typos, dmtn-135 talbe update

Author: womullan

RTN-003.tex

@@ -21,7 +21,7 @@
 
 
 \setDocRef{RTN-003}
-\setDocUpstreamLocation{\url{https://github.com/rubin-observatory/rtn-003}}
+\setDocUpstreamLocation{\url{https://github.com/lsst/rtn-003/}}
 
 \date{\vcsDate}
 
@@ -40,6 +40,7 @@
   \addtohist{1}{2020-06-22}{Migrated document from LPM-251}{Leanne Guy}
   \addtohist{1.1}{2020-08-05}{Fix some inconsistencies }{WOM}
   \addtohist{1.2}{2020-10-20}{Fix some more inconsistencies }{WOM}
+  \addtohist{1.3}{2021-06-29}{Rubin template, minor typos, update dmtn-135 tables }{WOM}
 }
 
 

cost.tex

@@ -6,7 +6,7 @@ \subsection{Maximizing Profits with Science-Driven IDACs}
 There are two main cost impacts of IDACs being set up outside of the \gls{US} and Chilean DACs: the positive impact is that some computational load may be taken off of these existing DACs, but the negative impact is the level of support required from the \RO Project in order to get them set up and running. This negative impact could be mitigated by ensuring that science productivity is maximized as a result of this extended effort. One way to do this might be to associate specific areas of science to a given \gls{IDAC}, and encourage users working in that field to use that \gls{IDAC}. This could create a customer base for the \gls{IDAC}, bring together like-minded experts, and effectively distribute the computing load across a network of IDACs. This might also enhance internal funding arguments for investment resources by arguing for synergies with local science goals and attracting international users and official endorsement.
 
 \subsection{Data Transfer}\label{sec:xfer}
-Even with good networks the data transfer will not be trivial, and could be quite expensive. \RO is not currently set up to distribute data to multiple sites, i.e., there is no form of peer-to-peer sharing. The bandwidth at \gls{NCSA} is adequate for receiving data and delivering {\tt Alerts} to brokers during the night; perhaps some day time bandwidth could be used to transfer data to IDACs. A full data release of images and catalogs does not have to transferred within a given day; if the correct agreements are in place with an \gls{IDAC}, a full release could be transferred slowly as it is produced, and then made available to the IDACs users in whole on the official release day.
+Even with good networks the data transfer will not be trivial, and could be quite expensive. \RO is not currently set up to distribute data to multiple sites, i.e., there is no form of peer-to-peer sharing. The bandwidth at \gls{USDF} is adequate for receiving data and delivering {\tt Alerts} to brokers during the night; perhaps some day time bandwidth could be used to transfer data to IDACs. A full data release of images and catalogs does not have to transferred within a given day; if the correct agreements are in place with an \gls{IDAC}, a full release could be transferred slowly as it is produced, and then made available to the IDACs users in whole on the official release day.
 \input{xfercost}
 
 \subsection{Compute {\it vs.} Storage Resources} \label{sec:cvs}

dmtn-135

@@ -4,7 +4,7 @@ \section{Introduction}\label{sec:intro}
 The current model for \VRO is that is provides proprietary data to approved users in Chile and the \gls{US}. The data access model accommodates this restricted data rights policy. This policy requires control over access, publication, and sharing of proprietary data which any (\gls{IDAC}) would have to comply with just as the \gls{US} and Chile DACs do.
 
 Access to \RO data products for any users will be possible through a \gls{DAC}. The United States's \gls{DAC} , referred to as the \gls{US} Data Facility, is
-where registered \RO users will perform scientific queries. Most users will have access to a default set of resources at the \gls{DAC} sufficient for basic queries and analysis. Users who require more resources will be able to apply for them, and those granted additional resources will be allowed (for example) to perform analysis on the full data releases using the \gls{RSP}. The \gls{RSP} is documented with the vision given in \citeds{LSE-319}, with more formal requirements in \citeds{LDM-554} and the design in \citeds{LDM-542}. The Chilean \gls{DAC} will be equivalent in functionality to the \gls{US} \gls{DAC}, but scaled-down in terms of the computational resources available for query and analysis given the smaller Chilean community \citedsp{LDM-572}.
+where registered \RO~ users will perform scientific queries. Most users will have access to a default set of resources at the \gls{DAC} sufficient for basic queries and analysis. Users who require more resources will be able to apply for them, and those granted additional resources will be allowed (for example) to perform analysis on the full data releases using the \gls{RSP}. The \gls{RSP} is documented with the vision given in \citeds{LSE-319}, with more formal requirements in \citeds{LDM-554} and the design in \citeds{LDM-542}. The Chilean \gls{DAC} will be equivalent in functionality to the \gls{US} \gls{DAC}, but scaled-down in terms of the computational resources available for query and analysis given the smaller Chilean community \citedsp{LDM-572}.
 
 %This document proposes a set of guidelines and policies for partner institutions -- in the US, Chile, or one of the International Contributors with signed Memoranda of Agreement -- that are interested in hosting the LSST data, in whole or in part, for their affiliated members as an independent Data Access Center (IDAC).
 The following sections include the types of data products that could be hosted (Section \ref{sec:data}), the requirements and responsibilities that would be expected of an \gls{IDAC} hosting \RO proprietary data products (Section \ref{sec:reqs}), and a description of the main costs {\it vs.} their science impacts (Section \ref{sec:costs}).

intro.tex

@@ -16,7 +16,9 @@ \subsection{\RO site topology} \label{sec:topology}
 \end{figure}
 
 \subsection{Authentication and Access}\label{sec:auth}
-Any DAC will have to support authentication according to \RO access rules. This may imply delegating access control to a \RO authorization service. In addition any user with access rights should be allowed access tot he IDAC.
+Any DAC will have to support authentication according to \RO access rules. This may imply delegating access control to a \RO authorization service. In addition any user with access rights should be allowed access to the IDAC.
+
+We should be clear that on the construction side we have not necessarily planned to Authenticate users at various IDACs so there may be some development work needed to make an Auth service available.
 
 \subsection{Required Resources} \label{sec:resources}
 Institutions or organizations wishing to set up independent data access centres will be expected to have
@@ -66,4 +68,6 @@ \subsection{Data Distribution} \label{sec:dist}
 
 In \gls{HEP} experiments such as  BaBar various physics analysis groups (science collaborations in \gls{LSST} ) were assigned to specific international centers as their primary computing and analysis facility, thereby distributing the computing load around the "network." Users naturally tend to use the facility with available resources and cycles.
 
+As of 2021 the baseline for large transfers looks like Rucio\footnote{\url{https://rucio.cern.ch/}}.
+