Agenda

• Running example (see below)
• Need to assign actions to assess/become familiar with vocabularies that might be good to re-use, e.g. those now listed on PmArchitecturalDirectionV2

Minutes

• Decided to model Top Processes in terms of CPU, Memory as rdf:Seq collections
• JohnArwe suspects for these small lists the benefits outweigh any issues around indexing, but will get a second opinion from ArthurRyman.

Examples

• Updates to the Running Example showing various alternatives. Read each step as an add-on to the previous.
  • Adding back in content (pm:MetricDefinition, pm:PerfMonRec) that was lost in the 6/28 to 7/10 re-cast Turtle RDF/XML
  • Assign all blank nodes a hash URI (so they are no longer blank nodes) Turtle RDF/XML
  • When a process occurs in >1 list, simply create another process-info resource for each list it occurs in. The name and process-level metrics are repeated, but there is only 1 ranking needed per process info entry. I did not do this, it's pretty obvious that you just need additional nodes (blank or hash-URL-identified or server-addressable).
  • Use of rdf:List Turtle RDF/XML
  • Use of rdf:Seq Turtle RDF/XML
  • Use of W3C Submission Containers, one per process list Turtle RDF/XML

Issues

Actions

What other vocabularies should we be looking at (besides QUDT, EMS, CRTV). Need something for describing statistics for example? Owner:TBD

Need to continue filling out predicates for non-ComputerSystem resources? Virtual Servers - Janet, Database/Database Servers - Julie, Web App Server, http Servers - Julie/Ravi

Need a recommendation for metric meta deta resource. Owner: TBD