[Sea-sa] Meetings notes for SEA - SA, RASDS/RASIM Subset meeting, 12 August 24

[Shames, Peter M (US 312B)]

Dear SEA SAWG team,

Attendees: Josiah Johnson, Shelbun Cheng, Fred Slane, Paul Thompson, David Giaretta, Peter Shames

Great meeting Monday, 12 Aug 24, so glad to have your participation.  The focus was largely on the proposed new work item, the refresh of the Reference Architecture for Space Information Management (RASIM), CCSDS 312.0-G-1.

Further work on RASDSv2 will be delayed until after the results of the Agency review have been returned.

Feedback and corrections on these notes is hereby solicited from those who were present.

General notes from the telecon:

DAI WG Discussion

  *   David Giaretta, MOIMS DAI WG Chair, introduced the WG to the fundamental concepts of their new archive interoperability framework.  This is intended as a formal, interoperable, Blue Book specification for implementing instances of the OAIS conceptual archive architecture.
  *   This OAIS Architecture document, and related test reports (Yellow Books), are in development now.
  *   This implementation architecture adopts HTTP/REST at the messaging layer (over TCP/IP) and adopts JSON at the data representation and exchange layer.
  *   David agreed that the new BB is largely aligned with the existing RASIM, and the planned updates.  There appear to have been some changes in the DAI use of terminology related to data, information, and metadata. We had a somewhat lengthy discussion of these differences, and the matter is not yet settled.
  *   Peter Shames briefly presented some of the RASDS Information Viewpoint materials that directly address these topics (RASDSv2, fig 10-4), and also described the relationships that are defined between abstract architecture, data models, and actual data objects as realized in a system (RASDSv2, fig 10-5).
  *   RASIM is an abstract architecture and it defines terms, functions, and relationships along with a variety of potential deployment options.   The OAIS Architecture is an actual deployment architecture with defined interfaces and protocols.  These are defined at different stages of specificity in the context of RASDS Fig 10-5.
  *   Some of the disconnects that appear to exist between RASDSv2 and the OAIS architecture may stem from a lack of alignment of terminology usage between the two WGs.
  *   There was general agreement that terminology had to be either defined clearly, and aligned or resolved in some way.
  *   Fred Slane (wearing his ISO TC20/SC14 hat) suggested that there might be interest in that organization to use some of this implementation framework to support the On Orbit Servicing (OOS) work that CONFERS is doing (https://satelliteconfers.org/).
RASIM Discussion

  *   Josiah Johnson presented the draft set of RASIM document revisions that he has been editing.
  *   The RASIM is fundamentally a Green Book, a descriptive document, but it contains accessible definitions of a lot of abstract information system elements, such as servers, registries, repositories, query and products servers, and how they may be accessed and used to design information management systems.
  *   So, while this document is a GB, it has a lot of materials that might appear in a more formal Magenta Book.  The WG may consider changing the nature of this document.
  *   Josiah has added a new section that addresses implementation and deployment considerations for information systems.
  *   When RASIM was first published most of the information systems were fairly straightforward instances of client / server architectures.
  *   Modern systems, driven by the explosion of available information and higher powered process and product approaches have adopted cloud-based, containerized, and other hybrid deployment models.  Josiah has provided a start of a set of clear descriptions of the distinctions among these different deployment approaches.
  *   What came out of this discussion was a list of topics that need to be addressed for each option:
     *   Options presented using a consistent style, adopting RASDS viewpoints and representations as needed
     *   Providing pros and cons of each option
     *   Parallel discussion for each option of considerations such as: security, complexity, cost, development / deployment timeline
     *   Consideration of whether some comparison table would also be a benefit to readers
  *   We had a discussion of the distinction between “thick” vs “thin” clients, and of the option that now exists of “thin client” boxes, that are really inexpensive client-side, programmable, processors (which probably should be called something like “svelte clients”).  These are neither thin nor thick, but in between.
  *   Josiah will be making updates to what was discussed, and that draft will be shared with the WG on the CWE.
Definitions

  *   An observation from both the OAIS discussion and the RASIM discussion is that clear definitions of terms is essential.  CCSD common practice is to use terms already defined in the SANA by preference, wherever that is possible (https://sanaregistry.org/r/terms/).
  *   New terms can be defined, and often are where clear definitions are not available, but care should be taken to first seek as “legitimate” a source of definitions as possible.  In addition to SANA this might include: NIST, ISO, IEEE, or other on-line accessible sources.
  *   If more than one definition seems viable care must be taken to do disambiguation and to ensure alignment across the terminology set.
  *   This will become necessary in the DAI WG / SAWG discussion, as well as internal to RASIM and within the SAWG.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.ccsds.org/pipermail/sea-sa/attachments/20240812/92096e9e/attachment-0001.htm>