Thank you, you answered all my questions. We want to try if we can improve performance of the Apache ODE BPEL engine by copying variables (which you do not have) and xpath evaluation within the database engine. And we would like to compare our approach with what is done in other workflow systems, since there is very little to no information in the papers about the workflow engines on that particular topic I try to mail to the dev lists to get some. However, if you are interested in the results I can send you a copy of my work when it's finished or post a link to this mailing list.
Regards, Florian Am 04.10.2010 11:21, schrieb Stian Soiland-Reyes: > On Mon, Oct 4, 2010 at 08:20, Florian Wagner - FloWSoft.de > <[email protected]> wrote: >> I currently writing my diploma thesis "exploit an internal database for >> an improved workflow execution" and would like to ask you, if you are >> exploiting your internal database in any kind to improve the performance >> of SCUFL workflows. > Taverna's internal database is used for storing intermediate data and > the provenance of what has happened in a workflow run. > > Using an internal database allow us to process many large pieces of > data that would not all fit in memory at the same time. In that way > the database is used almost like a swap file. > > Additionally the results and intermediate values of workflows run with > provenance on and database storage on ('In-memory' ticked off) will > also be available for inspection on later runs of Taverna. > >> I am aware of the fact that this is a data centric workflow model and >> that you mainly store the intermediate and end results to show them in >> the result explorer and overgive them to the next processor. But if you >> do something else than storing and loading data with your database it >> would be nice to know. > We don't "exploit" the database in any way other than storing and > retrieving values. I guess with exploiting here you are thinking about > generating clever queries to perform parts of the workflow or similar. > > We do have automatic query generation for performing provenance > lookups. Using the Provenance API a developer can ask questions like > 'Which partsof a workflow was involved in generating value X' - which > will build a dynamic SQL query to retrieve the answer, based on the > recorded information and the static structure of the workflow, also > recorded in the provenance-side of the database. > ------------------------------------------------------------------------------ Virtualization is moving to the mainstream and overtaking non-virtualized environment for deploying applications. Does it make network security easier or more difficult to achieve? Read this whitepaper to separate the two and get a better understanding. http://p.sf.net/sfu/hp-phase2-d2d _______________________________________________ taverna-hackers mailing list [email protected] Web site: http://www.taverna.org.uk Mailing lists: http://www.taverna.org.uk/about/contact-us/ Developers Guide: http://www.taverna.org.uk/developers/
