Comparison of Hibernate with Derby server vs DataNucleus with H2 server

Each of the following tables focuses on a specific database operation, where the last table presents average results comparison.

Speed comparison of JPA database persistence operations (normalized score, higher is better)

Transaction Size =>	Few Entities		Many Entities		Average Score
	Hibernate Derby server	DataNucleus H2 server	Hibernate Derby server	DataNucleus H2 server	Hibernate Derby server	DataNucleus H2 server
Basic Person Test	3.8	3.4	3.0	2.2	3.4	2.8
Element Collection Test	2.0	1.2	1.3	0.64	1.7	0.92
Inheritance Test	3.4	3.0	3.0	1.9	3.2	2.4
Indexing Test	4.8	5.3	4.1	3.5	4.4	4.4
Graph (Binary Tree) Test	1.8	2.1	1.6	1.3	1.7	1.7
Multithreading Test	6.9	10.0	3.9	4.3	5.4	7.2
All Tests	3.8	4.2	2.8	2.3	3.3	3.2

The results above show that in general Hibernate with Derby server is slightly more efficient than DataNucleus with H2 server in persisting JPA entity objects to the database.

A large performance gap has been detected when using JPA element collections with large transaction size. Comparing the normalized speed of DataNucleus with H2 database server (0.64) to the normalized speed of Hibernate with Derby database server (1.3) reveals that in that case, Hibernate with Derby server is 2.0 times faster than DataNucleus with H2 server.

Speed comparison of JPA database retrieval operations (normalized score, higher is better)

Retrieval Size =>	Few Entities		Many Entities		Average Score
	Hibernate Derby server	DataNucleus H2 server	Hibernate Derby server	DataNucleus H2 server	Hibernate Derby server	DataNucleus H2 server
Basic Person Test	3.8	2.5	8.4	5.2	6.1	3.9
Element Collection Test	1.0	0.56	1.0	1.1	1.0	0.82
Inheritance Test	3.5	2.2	9.7	6.1	6.6	4.2
Indexing Test	3.2	2.1	9.5	7.7	6.4	4.9
Graph (Binary Tree) Test	0.91	5.7	1.2	8.9	1.1	7.3
Multithreading Test	7.8	2.3	13.4	8.7	10.6	5.5
All Tests	3.4	2.5	7.2	6.3	5.3	4.4

The results above show that in general Hibernate with Derby server is slightly more efficient than DataNucleus with H2 server in retrieving JPA entity objects from the database.

A large performance gap has been detected when using multithreading with small retrieval size. Comparing the normalized speed of DataNucleus with H2 database server (2.3) to the normalized speed of Hibernate with Derby database server (7.8) reveals that in that case, Hibernate with Derby server is 3.4 times faster than DataNucleus with H2 server.

On the other hand, Hibernate with Derby server is slower, for instance, when using graphs of objects with large retrieval size. Comparing the normalized speed of Hibernate with Derby database server (1.2) to the normalized speed of DataNucleus with H2 database server (8.9) reveals that in that case, Hibernate with Derby server is 7.4 times slower than DataNucleus with H2 server.

Speed comparison of JPA database query operations (normalized score, higher is better)

Retrieval Size =>	Few Entities		Many Entities		Average Score
	Hibernate Derby server	DataNucleus H2 server	Hibernate Derby server	DataNucleus H2 server	Hibernate Derby server	DataNucleus H2 server
Basic Person Test	34.0	4.1	4.1	1.3	19.1	2.7
Element Collection Test	33.8	4.0	0.64	0.49	17.2	2.3
Inheritance Test	20.5	2.9	5.0	1.5	12.8	2.2
Indexing Test	2.0	0.0036	9.2	1.5	5.6	0.75
Multithreading Test	36.2	1.3	5.1	failed	20.6	1.3
All Tests	25.3	2.5	4.8	1.2	15.1	1.9

DataNucleus with H2 server has failed in 1 tests (see exceptions).

The results above show that in general Hibernate with Derby server is much more efficient than DataNucleus with H2 server in executing the tested JPA queries. Comparing the normalized speed of DataNucleus with H2 database server (1.9) to the normalized speed of Hibernate with Derby database server (15.1) reveals that in these tests, Hibernate with Derby server is 7.9 times faster than DataNucleus with H2 server.

A huge performance gap has been detected when using database indexes with small retrieval size. Comparing the normalized speed of DataNucleus with H2 database server (0.0036) to the normalized speed of Hibernate with Derby database server (2.0) reveals that in that case, Hibernate with Derby server is 556 times faster than DataNucleus with H2 server.

Speed comparison of JPA database update operations (normalized score, higher is better)

Transaction Size =>	Few Entities		Many Entities		Average Score
	Hibernate Derby server	DataNucleus H2 server	Hibernate Derby server	DataNucleus H2 server	Hibernate Derby server	DataNucleus H2 server
Basic Person Test	2.9	2.5	2.5	1.4	2.7	1.9
Element Collection Test	2.3	2.0	1.5	1.0	1.9	1.5
Inheritance Test	2.9	2.4	3.5	2.3	3.2	2.3
Indexing Test	3.0	2.7	4.2	3.0	3.6	2.9
Graph (Binary Tree) Test	1.2	1.8	0.58	0.60	0.91	1.2
Multithreading Test	6.8	failed	2.5	failed	4.6	failed
All Tests	3.2	2.3	2.5	1.7	2.8	2.0

DataNucleus with H2 server has failed in 2 tests (see exceptions).

The results above show that in general Hibernate with Derby server is more efficient than DataNucleus with H2 server in updating JPA entity objects in the database.

Speed comparison of JPA database removal operations (normalized score, higher is better)

Transaction Size =>	Few Entities		Many Entities		Average Score
	Hibernate Derby server	DataNucleus H2 server	Hibernate Derby server	DataNucleus H2 server	Hibernate Derby server	DataNucleus H2 server
Basic Person Test	2.9	3.4	3.5	1.3	3.2	2.3
Element Collection Test	1.7	1.5	0.87	0.68	1.3	1.1
Inheritance Test	2.9	2.7	3.6	1.3	3.2	2.0
Indexing Test	3.7	5.7	3.6	1.9	3.7	3.8
Graph (Binary Tree) Test	0.69	1.5	0.70	1.2	0.70	1.3
Multithreading Test	6.2	failed	5.1	5.1	5.6	5.1
All Tests	3.0	3.0	2.9	1.9	3.0	2.4

DataNucleus with H2 server has failed in 1 tests (see exceptions).

The results above show that in general Hibernate with Derby server is slightly more efficient than DataNucleus with H2 server in deleting JPA entity objects from the database.

A large performance gap has been detected when using class inheritance in the object model with large transaction size. Comparing the normalized speed of DataNucleus with H2 database server (1.3) to the normalized speed of Hibernate with Derby database server (3.6) reveals that in that case, Hibernate with Derby server is 2.8 times faster than DataNucleus with H2 server.

On the other hand, Hibernate with Derby server is slower, for instance, when using graphs of objects with small transaction size. Comparing the normalized speed of Hibernate with Derby database server (0.69) to the normalized speed of DataNucleus with H2 database server (1.5) reveals that in that case, Hibernate with Derby server is 2.2 times slower than DataNucleus with H2 server.

Comparison of JPA/Database speed - the averages (normalized score, higher is better)

Transaction/Retrieval Size	Few Entities		Many Entities		Average Score
	Hibernate Derby server	DataNucleus H2 server	Hibernate Derby server	DataNucleus H2 server	Hibernate Derby server	DataNucleus H2 server
Basic Person Test	9.5	3.2	4.3	2.3	6.9	2.7
Element Collection Test	8.2	1.9	1.1	0.78	4.6	1.3
Inheritance Test	6.6	2.6	5.0	2.6	5.8	2.6
Indexing Test	3.3	3.2	6.1	3.5	4.7	3.3
Graph (Binary Tree) Test	1.2	2.8	1.0	3.0	1.1	2.9
Multithreading Test	12.8	4.5	6.0	6.0	9.4	5.3
All Tests	7.1	2.9	4.0	2.8	5.6	2.9

The results above show that in general Hibernate with Derby server is more efficient than DataNucleus with H2 server in performing JPA database operations.

A large performance gap has been detected when using JPA element collections with small transaction/retrieval size. Comparing the normalized speed of DataNucleus with H2 database server (1.9) to the normalized speed of Hibernate with Derby database server (8.2) reveals that in that case, Hibernate with Derby server is 4.3 times faster than DataNucleus with H2 server.

On the other hand, Hibernate with Derby server is slower, for instance, when using graphs of objects with large transaction/retrieval size. Comparing the normalized speed of Hibernate with Derby database server (1.0) to the normalized speed of DataNucleus with H2 database server (3.0) reveals that in that case, Hibernate with Derby server is 3.0 times slower than DataNucleus with H2 server.

JPA Performance Benchmark (JPAB)