db4o Developer Community

Yes, the PolePosition open benchmark: http://www.db4o.com/about/productinformation/benchmarks/http://www.polepos.org/ Best.

thank you for response, but it's method pointsIndex - it indexes only points I call these method from db4o tutorial class com.db4o.f1.Main public class Main {    public static void main(String[] args) throws Exception {//        ...

Hi, I'm doing some scalablity tests storing instances of class A (.NET2.0) using db4o 6. Class A has the following structure: class A: 3 properties (2 int, 1 string) and 1 List<B> - 10 instances of B in the collection. Class B: 4 properties (2 int, 2 string) and 1 List<C> and 1 List<D> - 20 instances of C and D in each ...

As a final result on large data sets:Creation of 12 indexes on 1 million unindexed objects took ~ 2 hours (2x slower than 5.4)Average performance of inserts with 12 indexes per object after 1 million of objects are present is 600-700 msec per transaction with 1 object insert.Turning flush off increases performance ~3x-4x timesCache height of 1-3 ...

Carl Rosenberger:You will see the real advantage of the new BTree indexes if you:(1) work against hard disc, not against memory(2) have a couple of more objects in the database than 1000 or 2000 like in your last tests(3) do single object commits ...  You are not really doing us a favour by posting your numbers and talking about ''one ...

Is suppose that difference of our test is that we are placing many objects with 12 fields indexed. That's a real stress for indexing subsystem, but we are positive that we'll have such objects. The problem as we can see it now is in hard usage of disk resources. Index search requires many data to be read. If you have many indexes - it requires ...

Hello.We have downloaded recently published version 5.7 with B-Tree indexes and made our common performance measurement tests.It took 14 hours to fill a DB with 1.1 million of sample object with several indexed fields of different type with average deviation of values ~ 1 million of different values. We are commiting every 1000 objects, so it's ...

I have just tried benchmarking db4o and found its performance to be severly lacking, I can't belive that its so slow so its got to be something I'm not doing right. Using a modified version of the pilot example from chapter one I created 100,000 objects. For all the pilot's I set up their points to = 1. I then wanted to run a report which add's ...