[Hibernate-JIRA] Created: (HHH-4953) Inconsistent behavior when fetching cached subclass entities from 2L
by Guenther Demetz (JIRA)
Inconsistent behavior when fetching cached subclass entities from 2L
--------------------------------------------------------------------
Key: HHH-4953
URL: http://opensource.atlassian.com/projects/hibernate/browse/HHH-4953
Project: Hibernate Core
Issue Type: Bug
Components: caching (L2)
Affects Versions: 3.5.0-CR-2, 3.5.0-CR-1, 3.5.0-Beta-4, 3.5.0-Beta-3, 3.5.0-Beta-2, 3.5.0.Beta-1, 3.3.1, 3.2.5
Environment: Hibernate3.5.0CR2,HSQLDB, EHCache as 2L-cache Impl
Reporter: Guenther Demetz
Attachments: Test2LCache.jar
Enabling the hibernate second-level cache is leading to inconsistent behavior when fetching cached subclass entities as opposed to fetching non-cached subclass entities.
For more informations, please see
https://forum.hibernate.org/viewtopic.php?f=1&t=1002880&p=2425931#p2425931
Testcase attached.
Cause:
IMHO the bug consists in the fact, that
persister.getRootEntityName() instead to persister.getEntityName() is passed as 3thrd parameter
to the constructor of CacheKey in most places where CacheKey is used, for example in DefaultLoadEventListener.java:
new CacheKey(
event.getEntityId(),
persister.getIdentifierType(),
persister.getRootEntityName(),
source.getEntityMode(),
source.getFactory()
In this way the real entity name get swallowed and the type-saveness is not guaranteed anymore.
Workaround:
Subclass2 tofind = null;
try {
tofind = (Subclass2) em.find(Subclass2.class,"identifier1");
}
catch (ClassCastException cc) {
// swallow , this is a workaround
}
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://opensource.atlassian.com/projects/hibernate/secure/Administrators....
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
11 years, 10 months
[Hibernate-JIRA] Created: (HHH-3854) Issue with greedy loading of associations (default-lazy=false )
by Amar Singh (JIRA)
Issue with greedy loading of associations (default-lazy=false )
---------------------------------------------------------------
Key: HHH-3854
URL: http://opensource.atlassian.com/projects/hibernate/browse/HHH-3854
Project: Hibernate Core
Issue Type: Bug
Components: envers
Environment: Hibernate - core 3.4.0-SNAPSHOT, Hibernate-envers (from trunk)
Reporter: Amar Singh
Attachments: greedy-loading.zip
I have an Entity GP which has a many to one association to entity CA.
<?xml version="1.0"?>
<!DOCTYPE hibernate-mapping PUBLIC "-//Hibernate/Hibernate Mapping DTD 3.0//EN"
"http://hibernate.sourceforge.net/hibernate-mapping-3.0.dtd">
<hibernate-mapping package="org.hibernate.envers.test.integration.onetoone.bidirectional" default-lazy="false">
<class name="BidirectionalRefIngPK" table="BIDIRECTIONAL_REFERENCING_ENTITY">
<id name="id" type="long" column="BIDIRECTIONAL_REFERENCING_ID" >
<generator class="native"/>
</id>
<property name="data"/>
<many-to-one name="reference" cascade="save-update" class="BidirectionalRefEdPK" column="BIDIRECTIONAL_REFERENCED_ID"/>
</class>
<class name="BidirectionalRefEdPK" table="BIDIRECTIONAL_REFERENCED_ENTITY">
<id name="longId" type="long" column="BIDIRECTIONAL_REFERENCED_ID">
<generator class="native"/>
</id>
<property name="data"/>
</class>
</hibernate-mapping>
With this when I load a previous version of BidirectionalRefIngPK, the loading mechanism tries to create a proxy for BidirectionalRefEdPK, but in my case lazy is false so an internal call to createProxy will return NULL hence .
This happens in ToOneIdMapper.java in method mapToEntityFromMap around line 82, which then internally calls for AbstractEntityTuplizer.java line 395 for getProxyFactory() ending in NULL pointer exception because proxy factory is null as mine is a greedy loading.
TO REPRODUCE
~~~~~~~~~~~~~
Please find a zipped folder attached. The zipped file has 4 files:
- GreedyLoading.java (actual testng test case)
- BidirectionalRefEdPK.java (Entity 1)
- BidirectionalRefIngPK.java (Entity 2)
- greedyloading.hbm.xml
Please place the three java files in its package (org.hibernate.envers.test.integration.onetoone.bidirectional), couldn't find the right package so created test case accordingly. Put the greedyloading.hbm.xml in src/test/resources.
Now when testng runs GreedyLoading.java you should see NPE.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://opensource.atlassian.com/projects/hibernate/secure/Administrators....
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
11 years, 10 months
[Hibernate-JIRA] Created: (HHH-2891) Query cache fails on many-to-many select
by Oleg Efimov (JIRA)
Query cache fails on many-to-many select
----------------------------------------
Key: HHH-2891
URL: http://opensource.atlassian.com/projects/hibernate/browse/HHH-2891
Project: Hibernate3
Issue Type: Bug
Components: query-hql
Affects Versions: 3.2.5
Reporter: Oleg Efimov
I have two entities, User and Group, with many-to-many relation:
-------------------------------------------------------------------------------------------------- User.hbm.xml
<?xml version="1.0"?>
<!DOCTYPE hibernate-mapping PUBLIC
"-//Hibernate/Hibernate Mapping DTD 3.0//EN"
"http://hibernate.sourceforge.net/hibernate-mapping-3.0.dtd">
<hibernate-mapping package="test.hibernate">
<class name="User" table="`USER`">
<id name="id" type="long" unsaved-value="null">
<column name="USER_ID" not-null="true"/>
<generator class="native"/>
</id>
<property name="login">
<column name="LOGIN" not-null="true" length="255"/>
</property>
<bag name="groups" table="USER_GROUP" lazy="false" cascade="none">
<key column="USER_ID"/>
<many-to-many class="Group" column="GROUP_ID"/>
</bag>
</class>
</hibernate-mapping>
-------------------------------------------------------------------------------------------------- Group.hbm.xml
<?xml version="1.0"?>
<!DOCTYPE hibernate-mapping PUBLIC
"-//Hibernate/Hibernate Mapping DTD 3.0//EN"
"http://hibernate.sourceforge.net/hibernate-mapping-3.0.dtd">
<hibernate-mapping package="test.hibernate">
<class name="Group" table="`GROUP`">
<id name="id" type="long" unsaved-value="null">
<column name="GROUP_ID" not-null="true"/>
<generator class="native"/>
</id>
<property name="name">
<column name="GROUPNAME" not-null="true" length="100"/>
</property>
<bag name="users" table="USER_GROUP" lazy="true" inverse="true" cascade="all-delete-orphan">
<key column="GROUP_ID"/>
<many-to-many class="User" column="USER_ID"/>
</bag>
</class>
<query name="Group.select.by.login" cacheable="true"><![CDATA[
select user.groups from User user where user.login=?
]]>
</query>
</hibernate-mapping>
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
test.hibernate.User and test.hibernate.Group classes are just simple beans.
In Group.hbm.xml there is a query called "Group.select.by.login", it's intended to be cached. So I add corresponding properties to Hibernate configuraition:
--------------------------------------------------------------------------------------------------
properties.put("hibernate.cache.provider_class", "org.hibernate.cache.EhCacheProvider");
properties.put("hibernate.cache.use_query_cache", "true");
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Configuration has no extra parameters.
Then I'm trying to execute the query two times:
-------------------------------------------------------------------------------------------------- test fragment
SessionFactory sessionFactory = configuration.buildSessionFactory();
Session session = null;
try {
session = sessionFactory.openSession();
Query query = session.getNamedQuery("Group.select.by.login");
List<Group> groups;
groups = query.setParameter(0, "admin").list(); // +++++++++++++++This list() returns list of two Group objects, ok +++++++++++++++
System.out.println("-----------------------------------------------------------------------------------");
for (Group group : groups) {
System.out.println(group.getName());
}
System.out.println("-----------------------------------------------------------------------------------");
groups = query.setParameter(0, "admin").list(); // +++++++++++++++This list() returns list of two nulls +++++++++++++++
System.out.println("-----------------------------------------------------------------------------------");
for (Group group : groups) {
System.out.println(group.getName());
}
System.out.println("-----------------------------------------------------------------------------------");
} finally {
if (session != null) {
session.close();
}
sessionFactory.close();
}
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
The first query returns list of, in my case, two groups. The second query returns list of two nulls instead of the same list.
I've investigated the problem, and found the following:
Query return value is "user.groups", and during query analysis (org.hibernate.hql.ast.QueryTranslatorImpl:160) query return type is defined as "org.hibernate.type.BagType(test.hibernate.User.groups)". This seems to be the problem, as during query put operation org.hibernate.CollectionType.disassemble is called (org.hibernate.cache.StandardQueryCache:80), which simply returns null as owner is null.
If this query is not a misuse, then I think, something is to be done at the time of analysis, so that return type be somewhat different.
Thanks.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://opensource.atlassian.com/projects/hibernate/secure/Administrators....
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
11 years, 10 months
[Hibernate-JIRA] Created: (HHH-4732) Update Teradata Dialect for Teradata 13.0
by David Repshas (JIRA)
Update Teradata Dialect for Teradata 13.0
-----------------------------------------
Key: HHH-4732
URL: http://opensource.atlassian.com/projects/hibernate/browse/HHH-4732
Project: Hibernate Core
Issue Type: Improvement
Components: core
Affects Versions: 3.3.2
Environment: Hibernate 3.3.2/ Teradata JDBC Driver 13.0, Teradata Database 13.0
Reporter: David Repshas
Attachments: teradata.zip
Would like to update Teradata Dialect to work with Teradata Database 13.0.
I've include the following files with changes as noted:
org.hibernate.dialect.dialect.java
- added include of java.sql.statement
- added code to support the removal of temporary
files created when using referential integrity
The three new methods are:
public boolean supportsDropPreProcess()
public String performDropPreProcess(Statement stmt, String dropSql) throws SQLException
public void performDropPostProcess(Statement stmt, String dropSql) throws SQLException
==================================
org.hibernate.dialect.TeradataDialect.java
- added new include files
- modified registerFunction calls for "current_time" and "current_date"
- put registerKeyword list in alphabetical order and added keywords:
column,map,password,summary, type, value and year
- changed "getAddColumnString()" to return "Add"
- added new methods:
public ViolatedConstraintNameExtracter getViolatedConstraintNameExtracter()
public String extractConstraintName(SQLException sqlex)
public boolean supportsNotNullUnique()
public boolean supportsExpectedLobUsagePattern()
public boolean supportsUnboundedLobLocatorMaterialization()
public boolean supportsDropPreProcess()
public String performDropPreProcess(Statement stmt, String dropSql) throws SQLException
public void performDropPostProcess(Statement stmt, String dropSql) throws SQLException
==================================
org.hibernate.tool.hbm2ddl.SchemaExport
- added hooks for new methods:
supportsDropPreProcess,performDropPreProcess and performDropPostProcess
==================================
Also made some changes to some of the files in testsuite:
=============================================
test/hql/astParserLoadingTest.java
- Add instance check of TeradataDialect to get correct behavior
=============================================
test/legacy/foobartest.java
- Add instance check of TeradataDialect to get correct behavior
- modified SQL: Teradata requires "<>" rather than "!="
=============================================
test/typeParameters\typeparametertest.java
-needed to disable these tests as Teradata locking was causing
a deadlock situation when they run
=============================================
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://opensource.atlassian.com/projects/hibernate/secure/Administrators....
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
11 years, 10 months
[Hibernate-JIRA] Created: (HHH-3795) Unexpected behaviour of saveOrUpdate() for associations with delete-orphan
by Fabian Dankof (JIRA)
Unexpected behaviour of saveOrUpdate() for associations with delete-orphan
--------------------------------------------------------------------------
Key: HHH-3795
URL: http://opensource.atlassian.com/projects/hibernate/browse/HHH-3795
Project: Hibernate Core
Issue Type: Improvement
Environment: Hibernate 3.2.6.GA, PostgreSQL 8.3
Reporter: Fabian Dankof
Hello there,
I think saveOrUpdate() behaves somewhat unexpected for associations with delete-orphan cascading style:
Prerequisite: Association is set to cascade="all,delete-orphan" and nullable=true
1) Within a single session (i.e. a transient object):
* Load parent
* Delete child
* saveOrUpdate()
Result: Child is deleted from DB, a DELETE statement was issued
2) Across multiple session (i.e. a detached object):
* Have parent with valid ID, but empty list/set of children
* saveOrUpdate()
Result: Child is not deleted from DB, an UPDATE statement is issued which sets FK of child to 'null'
3) like 2) but nullable is FALSE
Result: Nothing happens.
With merge() everything works as expected for detached and transient instances. I would either expect saveOrUpdate() to
completely disregard DELETE statements OR behave similar to merge(). At least the different results of case 1) and 2)
are pretty confusing.
Best regards,
Fabian
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://opensource.atlassian.com/projects/hibernate/secure/Administrators....
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
11 years, 10 months
[Hibernate-JIRA] Created: (HHH-3628) Hilo optimizer problem in case of multiple threads accessing the sequence table
by Montagnon Cyril (JIRA)
Hilo optimizer problem in case of multiple threads accessing the sequence table
-------------------------------------------------------------------------------
Key: HHH-3628
URL: http://opensource.atlassian.com/projects/hibernate/browse/HHH-3628
Project: Hibernate Core
Issue Type: Bug
Affects Versions: 3.2.6
Environment: Sybase
Reporter: Montagnon Cyril
If 2 (or more) threads access the table storing the ids, this optimizer won't work and will try to insert entities with twice the same idea.
The problem is the way the HiLoOptimizer class generates the id :
public synchronized Serializable generate(AccessCallback callback) {
if ( lastSourceValue < 0 ) {
lastSourceValue = callback.getNextValue();
while ( lastSourceValue <= 0 ) {
lastSourceValue = callback.getNextValue();
}
hiValue = ( lastSourceValue * incrementSize ) + 1;
value = hiValue - incrementSize;
}
else if ( value >= hiValue ) {
lastSourceValue = callback.getNextValue();
hiValue = ( lastSourceValue * incrementSize ) + 1;
}
return make( value++ );
}
In the 'else if' part, the 'value' variable isn't reaffected, which means the current thread will try to insert entities with an id that has already been used by another thread. The value should be reset with hiValue - incrementSize.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://opensource.atlassian.com/projects/hibernate/secure/Administrators....
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
11 years, 10 months