OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

soa-rm-ra message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: Re: [soa-rm-ra] Registries and Repositories


Danny,

The following material had been incorporated into the RA (TheArchitecture) 
Wiki under the Interacting with Services model, but I removed it following 
objection of the use of the term "metadata."  Nevertheless, I did try to 
characterize the difference between registries and repositories in that 
discussion (see below).

There is excellent material out there on this subject and a lot of it from 
our own colleagues within the RM and RA like Duane and Matt.  I would also 
encourage engaging Kathryn Berringer in this discussion since not only is 
she a voting member of this TC but she is also the chair of the ebXML 
Registry TC.  The JAXR (JSR-093) spec is also a very good resource.

Regards...

 - Jeff E.

Metadata in Review



Metadata is descriptive information about the meaning of other data, in 
other words, "data about data."  As stated earlier, the principal purpose of 
metadata is to facilitate and improve the retrieval and understanding of 
information and data, thus, metadata and metadata management are paramount 
in helping realize the goals of SOA.  Metadata can be defined in a variety 
of standard formats and vocabularies.  One of the more popular formats today 
for describing metadata is the Extensible Markup Language (XML).  Its 
popularity is due in large part to the fact that it is both human and 
machine readable.



There are a myriad of types of metadata.  Some example types of metadata 
include the following:

  a.. "Data about data" - Metadata for describing the structure, format and 
definitions of data assets (sometimes referred to as "structural metadata");
  b.. "Data about available data" - Metadata for data/content discovery 
contained in a data instance catalog or index (sometimes referred to as 
"discovery metadata");
  c.. "Data about available services" - Metadata for service directory 
(service oriented data assets);
  d.. "Data about people, organizations, resources, etc." - Metadata 
contained in data directory services.


In addition to these types of metadata, it's often useful to classify 
metadata according to its level of descriptiveness using the "pages" 
metaphor for looking up information based on classification level.  The most 
common metadata pages classifications include the following:

  a.. White Pages - Metadata used for resource discovery and include such 
definitions as resource identifiers ("owners"), names, locations, and 
interface definitions;
  b.. Yellow Pages - Metadata used to describe high level generic content 
and may include subjects, keywords, broad categorizations, community names, 
and categories, temporal and spatial constraints;
  c.. Brown Pages - Metadata used to describe granular ("semantic") content 
at the community level and may include ontologies, taxonomic vocabularies, 
thesauri, specialized data models, business/functional capabilities.  This 
kind of metadata is especially important in specialized communities of 
interest where community members often need to interact in an ad-hoc fashion 
but may not necessarily talk the same "business language."  Publishing such 
metadata helps information providers and consumers understand one another's 
capabilities and also enable third-parties to mediate among them as 
necessary, achieving free information flow that wouldn't be possible among 
stove-piped systems;
  d.. Green Pages - Metadata used to describe product or service access 
methods and may include specific data access technology and environment 
requirements, security or Quality of Protection (QoP) requirements, Quality 
of Service (QoS) attributes such as transport protocol details and Service 
Level Agreements (SLAs) between the data asset producers and consumers.


As with the many types and formats for metadata, there are a few key 
standards-based approaches for modeling metadata.  Some of the leading 
industry initiatives for modeling metadata include the Dublin Core Metadata 
Initiative (DCMI) and the Object Management Group's (OMG's) Meta-Object 
Facility (MOF) and XML Metadata Interchange (XMI) specifications.  This SOA 
reference architecture does not recommend nor prescribe one modeling 
standard over another but rather recommends that where such standards exist, 
they should be favored over custom or proprietary approaches.



To facilitate electronic information exchange and information- and 
data-level integration, metadata is generally captured in a metadata 
registry/repository, which is a system that is used to store and manage 
metadata.  Typically, a metadata registry/repository is a software system 
that uses a database to store and search data, document formats, definitions 
of data, and relationships among data.  Not all registries use a separate 
repository store such as a database management system.  Nevertheless, it is 
easier to think of the repository as the holder of submitted content and to 
think of the registry as a catalog (similar to a card catalog in a public 
library) that describes the submitted content in the repository.  Some 
publications go further by differentiating a metadata registry/repository 
from a metadata catalog/repository, the latter described as a system used to 
store instances of metadata associated with individual data assets, which 
can be used by search portals and other software applications to locate the 
data assets that are relevant to user queries.  For purposes of this SOA 
reference architecture, we simply use the term metadata registry/repository 
as the combined registry, catalog, and repository.



Best practice suggests that metadata registries and registries and 
repositories be established according to industry standards, for example, 
the ISO/IEC 11179 Specification - Part 3.  According to the ISO/IEC 11179 
Part 3, the structure of a metadata registry is specified in the form of a 
conceptual data model.  The metadata registry is used to keep information 
about data elements and associated concepts, such as "data element 
 concepts," "conceptual domains," and "value domains."  Generically, these 
are all referred to as "metadata items."  Such metadata are necessary to 
clearly describe record, analyze, classify and administer data.


----- Original Message ----- 
From: "Danny Thornton" <danny_thornton2@yahoo.com>
To: <soa-rm-ra@lists.oasis-open.org>
Sent: Sunday, December 03, 2006 9:31 PM
Subject: [soa-rm-ra] Registries and Repositories


>I would like to add a Registry/Repository discussion
> to the agenda for this Wednesday RM providing we have
> time.  This relates to the Visibility section of the
> RA.
>
> http://wiki.oasis-open.org/soa-rm/TheArchitecture/ServiceView/Visibility
>
> The RA currently defines the registry as containing
> links to Service Description artifacts and the
> repository is defined as storing those artifacts.
>
> Another view of the repository is that the repository
> is the storage location for the services, the system
> of record of the services.  One of the things stored
> in the repository is the configuration management of
> the SOA service code.  In this view, the SOA
> repository becomes somewhat internal to the
> organization.  The majority of the Service Description
> and its artifacts reside in the service registry.
>
> Other thoughts on this?
>
> Danny
>
> __________________________________________________
> Do You Yahoo!?
> Tired of spam?  Yahoo! Mail has the best spam protection around
> http://mail.yahoo.com 




[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]