I am working in a project using Spring, Spring Data JPA, Spring Security, Primefaces...
I was following this tutorial about dynamic datasource routing with spring.
In this tutorial, you can only achieve dynamic datasource switching between a pre-defined datasources.
Here is a snippet of my code :
springContext-jpa.xml
<bean id="dsCgWeb1" class="org.apache.commons.dbcp.BasicDataSource">
<property name="driverClassName" value="${jdbc.driverClassName.Cargest_web}"></property>
<property name="url" value="${jdbc.url.Cargest_web}"></property>
<property name="username" value="${jdbc.username.Cargest_web}"></property>
<property name="password" value="${jdbc.password.Cargest_web}"></property>
</bean>
<bean id="dsCgWeb2" class="org.apache.commons.dbcp.BasicDataSource">
// same properties, different values ..
</bean>
<!-- Generic Datasource [Default : dsCargestWeb1] -->
<bean id="dsCgWeb" class="com.cargest.custom.CargestRoutingDataSource">
<property name="targetDataSources">
<map>
<entry key="1" value-ref="dsCgWeb1" />
<entry key="2" value-ref="dsCgWeb2" />
</map>
</property>
<property name="defaultTargetDataSource" ref="dsCgWeb1" />
</bean>
What i want to do is to make the targetDataSources map dynamic same as its elements too.
In other words, i want to fetch a certain database table, use properties stored in that table to create my datasources then put them in a map like targetDataSources.
Is there a way to do this ?
Nothing in AbstractRoutingDataSource forces you to use a static map of DataSourceS. It is up to you to contruct a bean implementing Map<Object, Object>, where key is what you use to select the DataSource, and value is a DataSource or (by default) a String referencing a JNDI defined data source. You can even modify it dynamically since, as the map is stored in memory, AbstractRoutingDataSource does no caching.
I have no full example code. But here is what I can imagine. In a web application, you have one database per client, all with same structure - ok, it would be a strange design, say it is just for the example. At login time, the application creates the datasource for the client and stores it in a map indexed by sessionId - The map is a bean in root context named dataSources
#Autowired
#Qualifier("dataSources");
Map<String, DataSource> sources;
// I assume url, user and password have been found from connected user
// I use DriverManagerDataSource for the example because it is simple to setup
DataSource dataSource = new DriverManagerDataSource(url, user, password);
sources.put(request.getSession.getId(), dataSource);
You also need a session listener to cleanup dataSources in its destroy method
#Autowired
#Qualifier("dataSources");
Map<String, DataSource> sources;
public void sessionDestroyed(HttpSessionEvent se) {
// eventually cleanup the DataSource if appropriate (nothing to do for DriverManagerDataSource ...)
sources.remove(se.getSession.getId());
}
The routing datasource could be like :
public class SessionRoutingDataSource extends AbstractRoutingDataSource {
#Override
protected Object determineCurrentLookupKey() {
HttpServletRequest request = ((ServletRequestAttributes)
RequestContextHolder.getRequestAttributes()).getRequest();
return request.getSession().getId();
}
#Autowired
#Qualifier("dataSources")
public void setDataSources(Map<String, DataSource> dataSources) {
setTargetDataSources(dataSources);
}
I have not tested anything because it would be a lot of work to setting the different database, but I thing that it should be Ok. In real world there would not be a different data source per session but one per user with a count of session per user but as I said it is an over simplified example.
The datasource used by a thread might change from time to time.
Should pay attention to concurrency, applications might get concurrency issues in concurrent environment.
thread-bound AbstractRoutingDataSource sample
It can be achieved with AbstractRoutingDataSource and keeping the information in the thread-local Variable. Here is a beautiful working example you can refer to:
Multi-tenancy: Managing multiple datasources with Spring Data JPA
Related
I am working on a migration project which involves upgrading the platform to Spring 4 with MyBatis. In the legacy code, transactions are handled at a central locations wherein call to start/end transactions are spread across various classes like service class, helper class and DAO class.
I managed to convert all service classes to spring managed component and DAO classes to support MyBatis-spring API. Problem is my service class use several other classes to perform a function and those classes are all instantiated manually and used. Now if i start a transaction on service class methods and perform database transactions inside other helper or DAO classes which are not spring managed, my transaction handling doesn't work correctly. I have illustrated this problem in the below code. Could you tell what are the ways to acheive transaction handling without modifying the code?
Example :
package com.service;
#Service
class MyService {
#Transactional( propagation=Propagation.REQUIRED)
public void processRequest () {
HelperClass helper = new HelperClass();
helper.performOperation();
}
}
package com.helper;
// this class is not spring bean
class HelperClass {
// MyBatis mapper class
private EmployeeMapper mapper;
public HelperClass () {
mapper = // retrieve mapper class bean from spring context
}
public performOperation () {
// call to mapper class insert operation
// call to mapper class update operation
}
}
package com.dao;
#Component
interface EmployeeMapper {
// method definition to perform database operation
}
Spring configuration details:
<context:component-scan base-package="com" />
<bean id="dataSource" class="com.mchange.v2.c3p0.ComboPooledDataSource" destroy-method="close">
....
....
</bean>
<bean id="transactionManager" class="org.springframework.jdbc.datasource.DataSourceTransactionManager">
<property name="dataSource" ref="dataSource" />
</bean>
<tx:annotation-driven transaction-manager="transactionManager" mode="aspectj" />
<mybatis:scan base-package="com.dao" />
<bean id="sqlSessionFactory" class="org.mybatis.spring.SqlSessionFactoryBean">
<property name="dataSource" ref="dataSource" />
<property name="mapperLocations"
value="classpath*:mybatis/*.xml" />
</bean>
In the above code HelperClass.performOperation() method is doing 2 database operations (insert,update). Say if insert succeeds and update fails, my database transaction doesn't get rollback. Since I already started the transaction at MyService.processRequest() should this not rollback the operations that are carried inside that method call? Correct me if my understanding is wrong.
Note: See my own answer to this question for an example of how I solved this issue.
I am getting the following exception in my Spring MVC 4 + Hibernate 4 project:
org.hibernate.LazyInitializationException: failed to lazily initialize
a collection of role: com.mysite.Company.acknowledgements, could not
initialize proxy - no Session
After reading a lot of other questions about this issue, I understand why this exception occurs, but I am not sure how to fix it in a good way. I am doing the following:
My Spring MVC controller calls a method in a service
The service method invokes a method in a DAO class
The method in the DAO class fetches an entity object through Hibernate and returns it to the calling service
The service returns the fetched object to the controller
The controller passes the object to the view (JSP)
The view tries to iterate on a many-to-many association that is lazily loaded (and is thus a proxy object)
The exception is thrown because the session is closed at this point, and the proxy cannot load the associated data
I have previously worked with PHP and doctrine2, and this way of doing things caused no problems. I am trying to figure out the best way to solve this, because the solutions that I found so far don't seem so great:
Load the association eagerly, potentially loading lots of unnecessary data
Call Hibernate.initialize(myObject.getAssociation()); - this means that I have to loop through associations to initialize them (I guess), and just the fact that I have to do this, kind of makes the lazy loading less neat
Using a Spring filter to leave the session open in the views, but I doubt that this is a good thing to do?
I tried to use #Transactional in my service, but without luck. This makes sense because I am trying to access data that has not yet been loaded after my service method returns. Ideally I would like to be able to access any association from within my view. I guess the drawback of initializing the associations within my service is that I have to explicitly define which data I need - but this depends on the context (controller) in which the service is used. I am not sure if I can do this within my controller instead without losing the abstraction that the DBAL layer provides. I hope that makes sense. Either way, it would be great if I didn't have to always explicitly define which data I want to be available to my view, but just let the view do its thing. If that is not possible, then I am just looking for the most elegant solution to the problem.
Below is my code.
View
<%# page contentType="text/html;charset=UTF-8" language="java" %>
<%# taglib prefix="c" uri="http://java.sun.com/jsp/jstl/core" %>
<h1><c:out value="${company.name}" /> (ID: <c:out value="${company.id}" />)</h1>
<c:forEach var="acknowledgement" items="${company.acknowledgements}">
<p><c:out value="${acknowledgement.name}" /></p>
</c:forEach>
Controller
#Controller
public class ProfileController {
#Autowired
private CompanyService companyService;
#RequestMapping("/profile/view/{id}")
public String view(Model model, #PathVariable int id) {
Company company = this.companyService.get(id);
model.addAttribute("company", company);
return "viewCompanyProfile";
}
}
Service
#Service
public class CompanyServiceImpl implements CompanyService {
#Autowired
private CompanyDao companyDao;
#Override
public Company get(int id) {
return this.companyDao.get(id);
}
}
DAO
#Repository
#Transactional
public class CompanyDaoImpl implements CompanyDao {
#Autowired
private SessionFactory sessionFactory;
#Override
public Company get(int id) {
return (Company) this.sessionFactory.getCurrentSession().get(Company.class, id);
}
}
Company entity
#Entity
#Table(name = "company")
public class Company {
#Id
#GeneratedValue(strategy = GenerationType.AUTO)
private int id;
// Other fields here
#ManyToMany
#JoinTable(name = "company_acknowledgement", joinColumns = #JoinColumn(name = "company_id"), inverseJoinColumns = #JoinColumn(name = "acknowledgement_id"))
private Set<Acknowledgement> acknowledgements;
public Set<Acknowledgement> getAcknowledgements() {
return acknowledgements;
}
public void setAcknowledgements(Set<Acknowledgement> acknowledgements) {
this.acknowledgements = acknowledgements;
}
// Other getters and setters here
}
Acknowledgement entity
#Entity
public class Acknowledgement {
#Id
private int id;
// Other fields + getters and setters here
}
mvc-dispatcher-servlet.xml (a part of it)
<bean id="sessionFactory" class="org.springframework.orm.hibernate4.LocalSessionFactoryBean">
<property name="dataSource" ref="dataSource" />
<property name="packagesToScan">
<list>
<value>com.mysite.company.entity</value>
</list>
</property>
<property name="hibernateProperties">
<props>
<prop key="hibernate.dialect">org.hibernate.dialect.PostgreSQL9Dialect</prop>
</props>
</property>
</bean>
<bean id="transactionManager" class="org.springframework.jdbc.datasource.DataSourceTransactionManager">
<property name="dataSource" ref="dataSource" />
</bean>
<tx:annotation-driven />
Thanks in advance!
As #Predrag Maric said writing service methods with different initialization for entity assotiations might be an option.
OpenSessionInView is a quit discussable pattern, which can be a solution in your case, though.
Another option is to set
<prop key="hibernate.enable_lazy_load_no_trans">true</prop>
property. It is designed to solve org.hibernate.LazyInitializationException problem and is available since hibernate 4.1.6.
So what does this property do? It simply signals Hibernate that it
should open a new session if session which is set inside current
un-initialised proxy is closed. You should be aware that if you have
any other open session which is also used to manage current
transaction, this newly opened session will be different and it may
not participate into the current transaction unless it is JTA. Because
of this TX side effect you should be careful against possible side
effects in your system.
Find more information here: http://blog.harezmi.com.tr/hibernates-new-feature-for-overcoming-frustrating-lazyinitializationexceptions and
Solve Hibernate Lazy-Init issue with hibernate.enable_lazy_load_no_trans
The simplest and most transparent solution is the OSIV pattern. As I'm sure you aware, there is plenty of discussion about this (anti)pattern and the alternatives both on this site and elsewhere so no need to go over that again. For example:
Why is Hibernate Open Session in View considered a bad practice?
However, in my opinion, not all criticisms of OSIV are entirely accurate (e.g. the transaction will not commit till your view is rendered? Really? If you are using the Spring implementation then https://stackoverflow.com/a/10664815/1356423)
Also, be aware that JPA 2.1 introduced the concept of Fetch Graphs which give you more control over what is loaded. I haven't tried it yet but maybe this if finally a solution for this long standing problem!
http://www.thoughts-on-java.org/2014/03/jpa-21-entity-graph-part-1-named-entity.html
For others who are looking for a solution, here is how I solved the problem.
As Alan Hay pointed out, JPA 2.1+ supports entity graphs, which ended up solving my problem. To make use of it, I change my project to use the javax.persistence.EntityManager class instead of the SessionFactory, which would then hand me the current session. Here is how I configured it in my dispatcher servlet configuration (some things excluded):
<beans xmlns="http://www.springframework.org/schema/beans"
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xmlns:context="http://www.springframework.org/schema/context"
xmlns:mvc="http://www.springframework.org/schema/mvc"
xmlns:jee="http://www.springframework.org/schema/jee"
xmlns:tx="http://www.springframework.org/schema/tx"
xsi:schemaLocation="http://www.springframework.org/schema/mvc http://www.springframework.org/schema/mvc/spring-mvc-4.1.xsd
http://www.springframework.org/schema/beans
http://www.springframework.org/schema/beans/spring-beans.xsd
http://www.springframework.org/schema/context http://www.springframework.org/schema/context/spring-context.xsd
http://www.springframework.org/schema/jee http://www.springframework.org/schema/jee/spring-jee-4.1.xsd
http://www.springframework.org/schema/tx http://www.springframework.org/schema/tx/spring-tx-4.1.xsd">
<jee:jndi-lookup id="dataSource" jndi-name="java:/comp/env/jdbc/postgres" expected-type="javax.sql.DataSource"/>
<bean id="entityManagerFactory" class="org.springframework.orm.jpa.LocalContainerEntityManagerFactoryBean">
<property name="dataSource" ref="dataSource" />
<property name="packagesToScan" value="dk.better.company.entity, dk.better.user.entity" />
<property name="jpaVendorAdapter">
<bean class="org.springframework.orm.jpa.vendor.HibernateJpaVendorAdapter" />
</property>
<property name="jpaProperties">
<props>
<prop key="hibernate.dialect">org.hibernate.dialect.PostgreSQL9Dialect</prop>
<prop key="hibernate.show_sql">true</prop>
</props>
</property>
</bean>
<bean id="transactionManager" class="org.springframework.orm.jpa.JpaTransactionManager">
<property name="entityManagerFactory" ref="entityManagerFactory" />
</bean>
<tx:annotation-driven />
</beans>
Below is an example of a DAO class.
#Transactional
public class CompanyDaoImpl implements CompanyDao {
#PersistenceContext
private EntityManager entityManager;
#Override
public Company get(int id) {
EntityGraph<Company> entityGraph = this.entityManager.createEntityGraph(Company.class);
entityGraph.addAttributeNodes("acknowledgements");
Map<String, Object> hints = new HashMap<String, Object>();
hints.put("javax.persistence.loadgraph", entityGraph);
return this.entityManager.find(Company.class, id, hints);
}
}
I found that if I did not use #PersistenceContext but Autowired instead, then the database session was not closed when rendering the view, and data updates were not reflected in subsequent queries.
For more information on entity graphs, I encourage you to read the following articles:
http://www.thoughts-on-java.org/2014/03/jpa-21-entity-graph-part-1-named-entity.html
http://www.thoughts-on-java.org/2014/04/jpa-21-entity-graph-part-2-define.html
My vote goes to Hibernate.initialize(myObject.getAssociation()) in service layer (which also means #Transactional should be moved from DAO to service methods)
IMHO, service methods should return all data that is required by the caller. If the caller wants to display some association on the view, it is service's responsibility to supply that data. This means you could have several methods that apparently do the same thing (return Company instance), but depending on the caller, different associations could be fetched.
On one project we had kind of configuration class, which contained information on what associations should be fetched, and we had a single service method which also accepted that class as a parameter. This approach meant we have only one method which is flexible enough to support all callers.
#Override
public Company get(int id, FetchConfig fc) {
Company result = this.companyDao.get(id);
if (fc.isFetchAssociation1()) {
Hibernate.initialize(result.getAssociation1());
}
...
return result;
}
Trigger in the service layer requires lazy loading of the Set's size method.
Tools:
public class HibernateUtil {
/**
* Lazy = true when the trigger size method is equal to lazy = false (load all attached)
*/
public static void triggerSize(Collection collection) {
if (collection != null) {
collection.size();
}
}
}
in your service method:
Apple apple = appleDao.findById('xxx');
HibernateUtil.triggerSize(apple.getColorSet());
return apple;
then use apple in controller, everything is ok!
I'm using Spring Web Flow to build an application. I am making use of the Flow Managed Persistence Context so the entity manager is 'kept open' during the execution of my flow and I can access lazy loaded properties (similar to OpenEntityManagerInViewFilter or OpenSessionInViewFilter for Spring MVC). When I use this, every time I submit a form, the number of active database connections increases, if I don't use the FMPC, I have no problems with the number of open connections).
I'm working with the following setup.
TransactionManager:
#Bean
#Autowired
public JpaTransactionManager transactionManager(EntityManagerFactory entityManagerFactory) {
return new JpaTransactionManager(entityManagerFactory);
}
DataSource:
#Bean
public DataSource dataSource() {
final BasicDataSource dataSource = new BasicDataSource();
dataSource.setDriverClassName(environment.getRequiredProperty(PROPERTY_DATABASE_DRIVER));
dataSource.setUrl(environment.getRequiredProperty(PROPERTY_DATABASE_URL));
dataSource.setUsername(environment.getProperty(PROPERTY_DATABASE_USERNAME, ""));
dataSource.setPassword(environment.getProperty(PROPERTY_DATABASE_PASSWORD, ""));
return dataSource;
}
EntityManagerFactory:
#Bean
public LocalContainerEntityManagerFactoryBean entityManagerFactory() {
final LocalContainerEntityManagerFactoryBean factoryBean = new LocalContainerEntityManagerFactoryBean();
factoryBean.setDataSource(dataSource());
factoryBean.setPackagesToScan(environment.getRequiredProperty(PROPERTY_ENTITYMANAGER_PACKAGES_TO_SCAN));
final JpaVendorAdapter vendorAdapter = new HibernateJpaVendorAdapter() {
{
setDatabase(Database.valueOf(environment.getRequiredProperty(PROPERTY_DATABASE_TYPE)));
setDatabasePlatform(environment.getRequiredProperty(PROPERTY_HIBERNATE_DIALECT));
}
};
factoryBean.setJpaVendorAdapter(vendorAdapter);
final Properties jpaProperties = new Properties();
jpaProperties.put(PROPERTY_HIBERNATE_FORMAT_SQL, environment.getRequiredProperty(PROPERTY_HIBERNATE_FORMAT_SQL));
jpaProperties.put(PROPERTY_HIBERNATE_NAMING_STRATEGY, environment.getRequiredProperty(PROPERTY_HIBERNATE_NAMING_STRATEGY));
jpaProperties.put(PROPERTY_HIBERNATE_SHOW_SQL, environment.getRequiredProperty(PROPERTY_HIBERNATE_SHOW_SQL));
jpaProperties.put(PROPERTY_HIBERNATE_HB2DDL_SQL, environment.getRequiredProperty(PROPERTY_HIBERNATE_HB2DDL_SQL));
factoryBean.setJpaProperties(jpaProperties);
return factoryBean;
}
JpaFlowExecutionListener:
#Bean
#Autowired
public JpaFlowExecutionListener jpaFlowExecutionListener(EntityManagerFactory entityManagerFactory, JpaTransactionManager transactionManager) {
return new JpaFlowExecutionListener(entityManagerFactory, transactionManager);
}
The BasicDataSource has maxActive set to 8 by default and when I reach 8 active connections, the page just hangs. Why are the connections not being closed after the request is complete? I have used the Chrome debugging tools (the network pane) to make sure there are not AJAX requests running or anything, my page submit (an HTTP POST) triggers a 301 redirect which then gives me a new HTTP GET and that results in a status 200, so all good.
When going from one page to the next, a service layer is called but as you can see from my beans, I am using the JpaTransactionManager and the SWF documentation says the following:
Note: All data access except for the final commit will, by default, be non-transactional. However, a flow may call into a transactional service layer to fetch objects during the conversation in the context of a read-only system transaction if the underlying JPA Transaction Manager supports this. Spring's JPA TransactionManager does support this when working with a Hibernate JPA provider, for example. In that case, Spring will handle setting the FlushMode to MANUAL to ensure any in-progress changes to managed persistent entities are not flushed, while reads of new objects occur transactionally.
For the sake of completeness, my spring-web-flow config:
<beans xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xmlns:webflow="http://www.springframework.org/schema/webflow-config"
xmlns="http://www.springframework.org/schema/beans"
xsi:schemaLocation="
http://www.springframework.org/schema/beans
http://www.springframework.org/schema/beans/spring-beans-3.0.xsd
http://www.springframework.org/schema/webflow-config
http://www.springframework.org/schema/webflow-config/spring-webflow-config.xsd">
<!-- Flow executor, repsonsible for creating and executing flows -->
<webflow:flow-executor id="flowExecutor" flow-registry="flowRegistry">
<webflow:flow-execution-listeners>
<webflow:listener ref="jpaFlowExecutionListener"/>
</webflow:flow-execution-listeners>
</webflow:flow-executor>
<!-- Flow registry, responsible for loading all flows so executor can execute them -->
<webflow:flow-registry id="flowRegistry" base-path="/WEB-INF/webflow/flows" flow-builder-services="flowBuilderServices">
<webflow:flow-location-pattern value="/**/*-flow.xml"/>
</webflow:flow-registry>
<!-- Flow builder services -->
<webflow:flow-builder-services id="flowBuilderServices" view-factory-creator="mvcViewFactoryCreator"/>
<!-- MvcViewFactoryCreator -->
<bean id="mvcViewFactoryCreator" class="org.springframework.webflow.mvc.builder.MvcViewFactoryCreator">
<property name="viewResolvers">
<list>
<ref bean="viewResolver"/>
</list>
</property>
</bean>
<!-- Flow handler adapter, responsible for answering request for a flow -->
<bean class="org.springframework.webflow.mvc.servlet.FlowHandlerAdapter">
<property name="flowExecutor" ref="flowExecutor"/>
</bean>
<!-- Flow handler mapping, lets Spring MVCs DispatcherServlet know to send flow request to SWF -->
<bean class="org.springframework.webflow.mvc.servlet.FlowHandlerMapping">
<property name="flowRegistry" ref="flowRegistry"/>
<property name="order" value="0"/>
<property name="interceptors">
<list>
<ref bean="localeChangeInterceptor" />
</list>
</property>
</bean>
</beans>
My flow has <persistence-context /> defined at the top.
I have the following end-state (which restarts the flow), even when I invoke this and the URL params change to e2s1, the number of active connections is not reset:
<end-state id="restart" commit="true" view="redirect:/main"/>
So it seems that the default hibernate property for hibernate.connection.release_mode is on_close. Considering the EntityManager is kept open during the whole flow, it never closes and a new connection is fetched from the pool for every request within the flow.
Changing the property to after_transaction solves this issue. However, in the case of fetching lazily loaded collections, it still doesn't work, each lazy property will fetch a new connection from the pool. In order to solve this I extended the JpaFlowExecutionListener with this:
public class AvoidLeakJpaFlowExecutionListener extends JpaFlowExecutionListener {
public AvoidLeakJpaFlowExecutionListener(EntityManagerFactory entityManagerFactory, PlatformTransactionManager transactionManager) {
super(entityManagerFactory, transactionManager);
}
#Override
public void paused(RequestContext context) {
super.paused(context);
EntityManager entityManager = (EntityManager) context.getFlowScope().get(PERSISTENCE_CONTEXT_ATTRIBUTE);
if (entityManager != null && entityManager instanceof HibernateEntityManager) {
HibernateEntityManager hibernateEntityManager = (HibernateEntityManager) entityManager;
hibernateEntityManager.getSession().disconnect();
}
}
}
This approach solves the lazily loaded collections problem but will still leak connections when loading of lazy-initialized entities is done using WebFlow's persistence context and this loading is performed during the transition to subflow that does not have configured. as described in in this bug report (where I found this solution as well).
I think I've read every question and answer on Spring and autowiring a servlet, both here and at springsource.org, and I still can't get it working.
All I want to do is have the datasource automatically set in my servlets. I understand that the container creates the servlet and not Spring.
Here is code from my test servlet:
package mypackage.servlets;
imports go here...
#Service
public class TestServlet extends HttpServlet
{
private JdbcTemplate _jt;
#Autowired
public void setDataSource(DataSource dataSource)
{
_jt = new JdbcTemplate(dataSource);
}
etc etc
In my applicationContext.xml I have:
<context:annotation-config />
<context:component-scan base-package="mypackage.servlets />
<import resource="datasource.xml" />
and in my datasource.xml:
<jee:jndi-lookup id="dataSource" jndi-name="java:comp/env/jdbc/db" />
If I can't get this working I'll just use WebApplicationContextUtils in the servlet's init method but I'd really like to make this work after all the reading I've been doing.
I'm using Spring 3, Java 1.6.
Thanks,
Paul
You need to replace your Servlets by Spring MVC contollers. Because Spring will not inject anything the classes (servlets) created by someone else then Spring itselfe (except #Configurable).
(To get an very simple example, take a look at the STS Spring Template Project: MVC).
What I wanted to do was get a DataSource reference in my Servlet for free, i.e. not calling a static getDatasource method on some class.
Here's what I learned and how I got it working:
Servlets cannot be configured or autowired by Spring. Servlets are created before Spring's app context is loaded. See issue SPR-7801: https://jira.springsource.org/browse/SPR-7801
What I did was create a DataSource in my applicationContext.xml and export that as a property:
<jee:jndi-lookup id="dataSource" jndi-name="java:comp/env/jdbc/db" />
<bean class="org.springframework.web.context.support.ServletContextAttributeExporter">
<property name="attributes">
<map>
<entry key="myDatasource">
<ref bean="dataSource"/>
</entry>
</map>
</property>
</bean>
In my servlet's init method I read the property:
public void init(ServletConfig config)
{
Object obj = config.getServletContext().getAttribute("myDatasource");
setDataSource((DataSource)obj);
}
public void setDataSource(DataSource datasource)
{
// do something here with datasource, like
// store it or make a JdbcTemplate out of it
}
If I'd been using DAOs instead of hitting the database from the servlets it would have been easy to wire them up for #Autowired by marking them #Configurable, and also be able to use #Transactional and other Spring goodies.
My application data access layer is built using Spring and EclipseLink and I am currently trying to implement the following feature - Ability to switch the current/active persistence unit dynamically for a user. I tried various options and finally ended up doing the following.
In the persistence.xml, declare multiple PUs. Create a class with as many EntityManagerFactory attributes as there are PUs defined. This will act as a factory and return the appropriate EntityManager based on my logic
public class MyEntityManagerFactory {
#PersistenceUnit(unitName="PU_1")
private EntityManagerFactory emf1;
#PersistenceUnit(unitName="PU_2")
private EntityManagerFactory emf2;
public EntityManager getEntityManager(int releaseId) {
// Logic goes here to return the appropriate entityManeger
}
}
My spring-beans xml looks like this..
<!-- First persistence unit -->
<bean class="org.springframework.orm.jpa.LocalContainerEntityManagerFactoryBean" id="emFactory1">
<property name="persistenceUnitName" value="PU_1" />
</bean>
<bean class="org.springframework.orm.jpa.JpaTransactionManager" id="transactionManager1">
<property name="entityManagerFactory" ref="emFactory1"/>
</bean>
<tx:annotation-driven transaction-manager="transactionManager1"/>
The above section is repeated for the second PU (with names like emFactory2, transactionManager2 etc).
I am a JPA newbie and I know that this is not the best solution. I appreciate any assistance in implementing this requirement in a better/elegant way!
Thanks!
First of all thanks to user332768 and bert. I tried using AbstractRoutingDataSource as mentioned in the link provided by bert, but got lost trying to hook up my jpa layer (eclipselink). I reverted to my older approach with some modifications. The solution looks cleaner (IMHO) and is working fine. (switching database at runtime and also writing to multiple databases in the same transaction)
public class MyEntityManagerFactoryImpl implements MyEntityManagerFactory, ApplicationContextAware {
private HashMap<String, EntityManagerFactory> emFactoryMap;
public EntityManager getEntityManager(String releaseId) {
return SharedEntityManagerCreator.createSharedEntityManager(emFactoryMap.get(releaseName));
}
#Override
public void setApplicationContext(ApplicationContext applicationContext)
throws BeansException {
Map<String, LocalContainerEntityManagerFactoryBean> emMap = applicationContext.getBeansOfType(LocalContainerEntityManagerFactoryBean.class);
Set<String> keys = emMap.keySet();
EntityManagerFactory entityManagerFactory = null;
String releaseId = null;
emFactoryMap = new HashMap<String, EntityManagerFactory>();
for (String key:keys) {
releaseId = key.split("_")[1];
entityManagerFactory = emMap.get(key).getObject();
emFactoryMap.put(releaseId, entityManagerFactory);
}
}
}
I now inject my DAO's with an instance (singleton) of MyEntityManagerFactoryImpl. The dao will then simply call createSharedEntityManager with the required release and will get the correct EntityManager for that database. (Note that i am now using application managed EntityManager and hence i have to explicitly close them in my dao)
I also moved to jta transaction manager (to manage transaction across multiple databases)
This is how my spring xml looks like now.
...
<bean class="org.springframework.orm.jpa.LocalContainerEntityManagerFactoryBean" id="em_Rel1">
<property name="persistenceUnitName" value="PU1" />
</bean>
<bean class="org.springframework.orm.jpa.LocalContainerEntityManagerFactoryBean" id="em_Rel2">
<property name="persistenceUnitName" value="PU2" />
</bean>
<bean class="org.springframework.transaction.jta.JtaTransactionManager" id="jtaTransactionManager">
</bean>
<tx:annotation-driven transaction-manager="jtaTransactionManager"/>
....
Cheers! (comments are welcome)
I am not sure if this is a clean method. Instead of declaring the enitiymanagerfactory multiple times, we can use the spring application context to get the entitymanagerfactory declared in the spring application.xml.
hm = applicationContext.getBeansOfType(org.springframework.orm.jpa.LocalEntityManagerFactoryBean.class);
EntityManagerFactory emf = ((org.springframework.orm.jpa.LocalEntityManagerFactoryBean) hm.get("&emf1")).getNativeEntityManagerFactory();
EntityManagerFactory emf2 = ((org.springframework.orm.jpa.LocalEntityManagerFactoryBean) hm.get("&emf2")).getNativeEntityManagerFactory();
This is something i need to do in the future too, for this i have bookmarked Spring DynamicDatasourceRouting
http://blog.springsource.com/2007/01/23/dynamic-datasource-routing/
As far as i understand, this is using one PU, which gets assigned different DataSources. Perhaps it is helpful.