Hi Damien, Thanks for the interest in Rya and for trying to include it in your study. I'm not sure why you have the problems. I cc-ed the Rya developers list (Rya is now an Apache incubator project) and hopefully someone has an idea.
Rya team - did anyone run into this kind of problems? Any suggestions? Thank you, Adina On Fri, Mar 25, 2016 at 9:55 AM, Damien Graux <[email protected]> wrote: > Hi, > > I am PhD student at Inria (France) and I am working with my team on SPARQL > query evaluators. More specifically, we are benchmarking a bunch of them > and we are trying to include RYA (see below for versions and system used). > > Currently, we are trying the run LUBM benchmark (as in " Rya: A Scalable > RDF Triple Store for the Clouds" 2012) but we have problems. We pinpointed > it to at least one specific problem, on some queries (e.g. Q14) RYA always > answer a list of empty elements ("[] [] [] …") whatever the dataset used > (we tried LUBM1, LUBM1k and LUBM10k). It seams that RYA gives the correct > number of answers but they are not supposed to be empty. > > We tried our best to solve this mystery but without success so we come to > your help so we can include RYA in upcoming our paper. Do you have an idea > where this might come from or can you redirect us towards someone in charge > of maintaining RYA? > > > We installed RYA 0.0.1 version : > commit e6be84a407e05c66cb4b4b6ef225d7e07dd10fcf > Author: Aaron Mihalik <[email protected]> <[email protected]> > Date: Fri Dec 4 20:25:51 2015 -0500 > > Rya is running on Cloudera 5.5.0 with Apache Accumulo 1.6. Everything runs > on a > cluster with the default configuration over 10 nodes (1 of primary & > secondary named nodes) > and each VM has 16 GB of ram, 4 TB of disk and 4 CPU (more than enough for > what is needed). > We did not activate the batch planner and followed the github installation > guide. > > Thank you in advance for any help you can provide, > > Warm Regards, > -- > Damien Graux > -- Dr. Adina Crainiceanu http://www.usna.edu/Users/cs/adina/
