From: Rob V. <rv...@do...> - 2013-05-09 20:11:58
|
Definitely I am looking into all of these and will get the appropriate changes made by end of tomorrow hopefully Re: speed tests I thing there may be an unintentional memory leak there since I've seen the same thing with the 1 million test on my local machine, need to do some digging so may just ignore those tests temporarily until such time as I've properly diagnosed the issue. Rob From: Tomasz Pluskiewicz <tom...@gm...> Reply-To: dotNetRDF Developer Discussion and Feature Request <dot...@li...> Date: Thursday, May 9, 2013 1:02 PM To: dotNetRDF Developer Discussion and Feature Request <dot...@li...> Subject: [dotNetRDF-Develop] Unit tests > Hi Rob > > I would like to discuss the remaining issue I've been having with running unit > tests both on my machine and in TeamCity. recent changes have corrected most > of the issues and now there are just a few remaining > > 1. DBpedia > If you look at the two latest builds on TC you will notice that out of the > blue some of the DBpedia tests would fail. Usually they report HTTP 500 and > weirdest is the graph equality test, which once in a while fetches a different > graph from the same URI. This has actually happened for me locally as well. > > I think we should replace all calls to DBpedia with out local copies of those > graphs and work on these just as most other tests do. > > 2. TriX performance tests > Not sure why these fail but this behavior has been consistent on TC for larger > test graphs. I guess it has something to do with mediocre performance of the > network drive that is likely acting for my VPS' storage. > > I think we could either exclude these tests from executing on CI environment > or maybe remove the timeout attribute and leave them for manual inspection. Or > maybe report as inconclusive if the parsing takes mroe than a given amount of > time? > > 3. Speed tests > I have completely disable the parsing speed tests because they were literally > killing my VPS. Memory usage that is. The 1 million triple test would kill JVM > on most runs. Locally I have 8 GB of RAM and upon completion this test uses > almost half of my physical memory. Is this really necessary or is there > possibly some memory leak? > > The remaining three tests are > > QueryTimeouts.SparqlQueryTimeoutDuringProductLazy2 > SparqlTests2.SparqlBindScope6 > SparqlQueryTests.SparqlFilterShouldGetPlaced > > Any ideas why these fail? > > Having the above sorted out I would like to start publishing NuGet from TC. > > Thanks, > Tom > ------------------------------------------------------------------------------ > Learn Graph Databases - Download FREE O'Reilly Book "Graph Databases" is the > definitive new guide to graph databases and their applications. This 200-page > book is written by three acclaimed leaders in the field. The early access > version is available now. Download your free book today! > http://p.sf.net/sfu/neotech_d2d_may___________________________________________ > ____ dotNetRDF-develop mailing list dot...@li... > https://lists.sourceforge.net/lists/listinfo/dotnetrdf-develop |