About reflection

forum_thread

Unexpected internal exception during set field of an Embeddable to null

Hello, we get an Unexpected internal exception and have no idea what goes wrong. We can reproduce (mostly) the error in a very big use case. In the big use case and we create and persist millions of SignalValueImpl objects (Embeddables). Partly as in the stack trace an Embeddable contain another Embeddable. It seems like that we can't reproduce the problem in smaller use cases with less objects. Stack trace: #1 2016-06-24 08:36 Hello, we get an Unexpected internal exception and have no idea what goes wrong. We can reproduce ...

 
manual

[ODB1] Chapter 5 - JDO Connections

The PersistenShows how to use database connections and transactions in JDO. ceManagerFactory interface represents a factory of database connections. Its main role is to provide PersistenceManager instances. The PersistenceManager interface represents a database connection. Every operation on a database requires a PersistenceManager instance. The Transaction interface represents a transaction on a database. Every operation that modifies the content of the database requires an active transaction.[toc hidden:1] This chapter describes three essential interfaces for using JDO: 5.1  javax.jdo.PersistenceManagerFactory 5.2  javax.jdo.PersistenceManager 5.3  javax.jdo.Transaction The PersistenceManagerFactory ...

 
release

2.5.5

... Fixed a bug in detecting orphan entities in reflection mode. Fixed a bug in listing inherited persistent ...

 
forum_thread

evictAll() behavior

I'm writing a multi-user drawing tool, using objectdb as the central store.  When a user press F5, I want to pick up all the changes from the database.  So, I call evictAll() in the following ways: pmf.getDataStoreCache().evictAll();  // the factory pm.evictAll();                                 // the persistence manager A couple of questions about this: 1. the first evictAll() seems to be clearing the L2 cache -- is this correct 2. the evicts only work when they are done inside a transaction -- is this specified in the JDO spec?... also related to a failure in using enhancement (since in reflection mode ObjectDB has to keep strong references to objects to detect ...

 
release

2.5.4

... Fixed a bug in detecting orphan entities in reflection mode. Fixed a bug in listing inherited persistent ...

 
release

2.5.3

... Fixed a bug in detecting orphan entities in reflection mode. Fixed a bug in listing inherited persistent ...

 
forum_thread

locks on pure query activities

analyzing the dynamic behavior we can see from the thread dump that a lot of our threads that are just doing simple queries show locks like these (question is: why there are locks and how could we get rid of that):   log1:   ... > <enhancement agent = "false" reflection = "warning" /> <cache ref = "weak" ...

 
forum_thread

Occasionally slow requests in load testing

We are running load testing against objectdb version 2.3.7_18. The performance is generally OK, however there are quite a few "spikes" in the performance where we are getting a lot of very slow response times from the DB, around 10 - 15 seconds. Looking at the DB server process with AppDynamics we can see that most of the requests get stuck in one of 2 places: com.objectdb.o.PGC:w: line number 232 com.objectdb.o.LFL:U: line number 1068 Please see the attached callgraph examples for full stack trace examples. ... > <enhancement agent = "false" reflection = "warning" /> <cache ref = "weak" ...

 
release

2.5.2

... Fixed a bug in detecting orphan entities in reflection mode. Fixed a bug in listing inherited persistent ...

 
release

2.5.1

... Fixed a bug in detecting orphan entities in reflection mode. Fixed a bug in listing inherited persistent ...