Cool! I'll give it a go on the weekend
thanks for the tips cheers Alex On Thu, Apr 24, 2008 at 5:43 AM, Jeremy Bauer <[EMAIL PROTECTED]> wrote: > Hi Alexander, > > You could try taking advantage of SQL statement batching by enabling it on > the dictionary. Batching is currently only enabled by default for DB2 and > Oracle providers but may work for postgres if the JDBC driver and > dictionary supports it. Check out the docs on statement batching here: > > http://openjpa.apache.org/builds/1.1.0-SNAPSHOT/docs/ref_guide_dbsetup_stmtbatch.html > > If it works, you'll likely need to do a bit of tuning to get the best > performance for your application. > > -Jeremy > > On 4/22/08, Alexander MacLennan <[EMAIL PROTECTED]> wrote: > > > > Hi, > > > > Its taking quite a long time in my code to store 3d model graphs to a > > postgresql 8.3 backed db > > > > I am generating large data sets.. so I would expect it not to be > > super-quick > > > > The data is being loaded in a loop, for each 3d model.. generate > features > > and store both the features and the model > > > > At the end of each loop iteration I persist the objects.. is there a > > smarter > > way of doing it so the commit speeds up? > > > > I have a Q9450, lots of memory and the db is sitting on a pair of > > (mirrored) > > 76GB WD Raptors ( there goes my scholarship money ). > > > > I'm using java6 > > > > cheers > > > > Alex > > > > > > On Fri, Apr 4, 2008 at 11:18 PM, Patrick Linskey <[EMAIL PROTECTED]> > > wrote: > > > > > Hi, > > > > > > Are you calling merge() or persist() a lot? There's a known > > > inefficiency in 1.1.0-SNAPSHOT (not 1.0) with merging right now. > > > > > > Also, what does the generated SQL look like in your examples? > > > > > > -Patrick > > > > > > On Fri, Apr 4, 2008 at 7:55 AM, Brill Pappin <[EMAIL PROTECTED]> wrote: > > > > We were using 1.0 in this project, but its showing up in another > > > > project using 1.1. > > > > > > > > I really *want* to use ojpa but I don't have time to mess around > and > > > > the problem is significant. > > > > There seem to be other inefficiencies as well that don't seem to > need > > > > to be there. > > > > > > > > Not sure what we're going to do about it though, all the ORM > > > > frameworks are going through some sort of growing pain at the > moment > > > > and I can't say I'm happy with any of them. > > > > > > > > - Brill Pappin > > > > > > > > > > > > > > > > On Fri, Apr 4, 2008 at 10:44 AM, Patrick T < > [EMAIL PROTECTED]> > > > wrote: > > > > > > > > > > My issue is even just fetching the entities. Despite of changing > > > > > relationships to LAZY fetch, I find that querying the DB takes a > > > really long > > > > > time. I have experienced the commit issue as well,but I've > worked > > > around > > > > > that by being careful on what I cascade through different > > entities. > > > > > > > > > > > > > > > > > > > > Brill Pappin wrote: > > > > > > > > > > > > As a rule you can expect *any* orm framework to be slower than > > > simple > > > > > > JDBC, > > > > > > there is simply more to do... What you gain from it is ease of > > > maintenance > > > > > > and implementation which will save you time later (often a lot > > of > > > time). > > > > > > As with anything, you need to weigh the benefits of any > solution > > > over > > > > > > another but in my experience most applications don't require > > such > > > high > > > > > > throughput that anyone would notice a couple of milliseconds > > saved > > > here > > > > > > and > > > > > > there... Of course it'll depend on what your doing :) > > > > > > > > > > > > However... > > > > > > > > > > > > I've just been doing some work tonight, and am finding that > > > begin() -> > > > > > > commit() transactions take longer and longer (in my case I > have > > a > > > very > > > > > > large > > > > > > update that is broken down into chunks). I haven't traced the > > > problem yet > > > > > > and it may be configuration related, however the time grows > > > exponentially. > > > > > > Does that sound like the same problem you are having? > > > > > > > > > > > > - Brill > > > > > > > > > > > > -----Original Message----- > > > > > > From: Patrick T [mailto:[EMAIL PROTECTED] > > > > > > Sent: Friday, March 28, 2008 3:55 PM > > > > > > To: [email protected] > > > > > > Subject: Slow JPA > > > > > > > > > > > > > > > > > > Hello Everyone, > > > > > > > > > > > > I'm just wondering if anyone else is finding JPA slower than > > > simple jdbc > > > > > > implementations. I have tried setting up some of the > attributes > > to > > > Lazy > > > > > > fetch and I still find it to be extremely slow. This is > causing > > my > > > web > > > > > > pages > > > > > > to load in the upwards of 10 seconds. I have tried running the > > > PCEnhancer > > > > > > and still takes a long time to fetch data. Does anyone have > any > > > > > > suggestions > > > > > > to speed it up? Thanks. > > > > > > > > > > > > Patrick > > > > > > -- > > > > > > View this message in context: > > > > > > http://www.nabble.com/Slow-JPA-tp16360845p16360845.html > > > > > > Sent from the OpenJPA Users mailing list archive at > Nabble.com. > > > > > > > > > > > > > > > > > > > > > > > > > > > > -- > > > > > View this message in context: > > > http://www.nabble.com/Slow-JPA-tp16360845p16491755.html > > > > > Sent from the OpenJPA Users mailing list archive at Nabble.com. > > > > > > > > > > > > > > > > > > > > > > > > > > -- > > > Patrick Linskey > > > 202 669 5907 > > > > > >
