Speaker
Description
In the past decade, Data and Analysis Preservation (DAP) has
gained an increased prominence in the scope of effort of major
High Energy and Nuclear Physics (HEP/NP) experiments, driven
by the policies of the funding agencies as well as realization
of the benefits brought by DAP to the science output of many
projects in the field. It is a complex domain which in addition to
archival of various tiers of data produced by the experiments also
includes long-term knowledge management as well as preservation
of relevant software and infrastructure elements in an organized and
functional state. In this presentation, we are sharing the experience,
technology choices and status of the effort of the PHENIX
Collaboration in this area for the first time.
The PHENIX Collaboration has recently undertaken an effort to put in
place Data and Analysis Preservation procedures and practices including
-
Use of Docker containers to preserve specialized and/or legacy
computing environments and enhance software portability -
Exploratory use of REANA as a mechanism to capture the
final stages of select analyses for preservation, validation
and training -
Active supervision and management of the materials created
and submitted by the Collaboration to the CERN HEPData portal
(Level 1 in standard DAP nomenclature), with a broad team
involvement -
Joining the CERN OpenData portal and using that platform to
host self-contained packages which include PHENIX special purpose
limited datasets and basic examples of analysis software
(Level 2 in standard DAP nomenclature) -
A vigorous team effort to migrate PHENIX research materials
from legacy information systems approaching end-of-life
to a robust and well maintained digital repository, opting to
use the Zenodo instance at CERN -
Deployment of a new Collaboration website for easy access to
curated materials including those obtained from legacy resources,
optimized for long-term stability and ease of maintenance
Common across all these work areas is the strategy of using
community-developed and supported tools, frameworks and services
while keeping in-house development to the absolute minimum. In summary,
this presentation covers
-
a real-life experience of a major experiment implementing a variety of DAP measures
against the background of legacy software and documentation systems -
an exposition of a variety of platforms and tools employed in PHENIX DAP
which can be used as a guidance in other HEP/NP projects
