Last week, I was busy with primarily with three activities.
First, I worked with Michael W to complete the design for the upload and query pages. He is working on the implementation of these two pages. As of now, the upload page would allow to upload trace files from various workflow systems. We have already developed a parser to parse Kepler/COMAD tarce files. In case, we need to load trace files from other workflow systems, we have to develop the respective parser for it. Our agile application design, based on the Abstract Factory Pattern (a core J2EE design pattern), allows us to add new parsers very easily. Michael W is very close to finish the implementations for these two pages. Michael W has run the PortPars workflow and generated the trace files.
Second, I have started developing a relational schema in MySQL for our gen_by/used provenance model. Last year, as part of the DataONE SoC, we developed a relational schema for “run” provenance. I am using that to build the new schema needed based on this year’s strategy. My plan is to be able to store trace data into both Neo4j and MySQL. The DAO layer would connect to the respective databases based on the database selection.
Third, as Michael A was back from his vacation, I explained him the changes made to our DAO layer so that he can incorporate the required changes in the Neo4j REST API. Once, this is done we would test our Provenace Repository using the web-based GUI we are developing.
Also, I developed a representative set of provenance queries for the gen_by/used provenance model, which should represent most of the provenance queries we know today. I would like extend the list as we mature our provenance model. I would finalize the design of our query page based on the final set of provenance queries.
Now, we are getting ready to test our application.