I have a follow-up: By default, Jetty starts Fuseki with -Xmx1200M for heap. Have you altered this for production? How many triples do you have and how often does your endpoint process queries? Our dataset won't be large at first (low millions of triples), but we can reasonably expect 10,000+ SPARQL queries per day. That's not a lot by dbpedia standards, but I have no idea how that compares to average LAM systems.
Thanks, Ethan On Thu, Feb 21, 2013 at 9:42 AM, Ethan Gruber <ewg4x...@gmail.com> wrote: > Thanks everyone for the info. This soothed my apprehensions of running > Fuseki in a production environment. > > Ethan > > > On Wed, Feb 20, 2013 at 4:05 PM, Ross Singer <rossfsin...@gmail.com>wrote: > >> I'll add that the LARQ plugin for Fuseki (which adds Lucene indexes) is >> pretty awesome, as well. >> >> -Ross. >> >> On Feb 20, 2013, at 3:57 PM, John Fereira <ja...@cornell.edu> wrote: >> >> > If forgot about that. That issue was created quite awhile ago and I >> hadn't check on it in a long time. I've found that Jetty has worked fine >> in our production environment so far. As I wrote earlier, I have it >> connecting to a jena SDB that is used for a semantic web application (VIVO) >> that was developed here. Although we have the semantic web application >> running on a different server than the SDB database I found the performance >> was fairly significantly improved by having the Fuseki server running on >> the same machine as the SDB. >> > >> > -----Original Message----- >> > From: Code for Libraries [mailto:CODE4LIB@LISTSERV.ND.EDU] On Behalf >> Of Ethan Gruber >> > Sent: Wednesday, February 20, 2013 2:52 PM >> > To: CODE4LIB@LISTSERV.ND.EDU >> > Subject: Re: [CODE4LIB] Fuseki and other SPARQL servers >> > >> > Hi Hugh, >> > >> > I have investigated the possibility of deploying Fuseki as a war in >> Tomcat ( >> > https://issues.apache.org/jira/browse/JENA-201) because I wasn't sure >> how the default Jetty container would respond in production, but since you >> aren't having any problems with that deployment, I may go ahead and do that. >> > >> > Ethan >> > >> > >> > On Wed, Feb 20, 2013 at 2:39 PM, Hugh Cayless <philomou...@gmail.com> >> wrote: >> > >> >> Hi Ethan! >> >> >> >> We've been using Jena/Fuseki in papyri.info for about a year now, >> iirc. >> >> We started with Mulgara, but switched. It's running in its own Jetty >> >> container in our system, but I've had no performance issues with it >> >> whatever. >> >> >> >> Best, >> >> Hugh >> >> >> >> On Feb 20, 2013, at 14:31 , Ethan Gruber <ewg4x...@gmail.com> wrote: >> >> >> >>> Hi all, >> >>> >> >>> I have been playing around with Fuseki ( >> >>> http://jena.apache.org/documentation/serving_data/index.html) for a >> >>> few months to get my feet wet with accessing and querying RDF. I >> >>> quite like it. I find it well documented and easy to set up. We >> >>> will soon deploy a SPARQL server in a production environment, and I >> >>> would like to know if others on the list have experience with Fuseki >> >>> in production, or have >> >> other >> >>> recommendations. Mulgara is off the table as it inexplicably >> >>> conflicts with other apps installed in Tomcat. >> >>> >> >>> Thanks, >> >>> Ethan >> >> >> > >
