jms.brokerurl=failover:(tcp://ip:61616?wireFormat.maxInactivityDuration=0)?jms.prefetchPolicy.queuePrefetch\=2&randomize\=false&initialReconnectDelay\=50&timeout\=3000
Spring configuration
<bean id="pooledConnectionFactory" class="org.apache.activemq.pool.PooledConnectionFactory" init-method="start" destroy-method="stop">
<property name="connectionFactory" ref="jmsConnectionFactory"/>
<property name="maxConnections" value="10"/>
<property name="maximumActive" value="10"/>
<property name="idleTimeout" value="0"/>
</bean>
<bean id="receiveContainer" class="org.springframework.jms.listener.DefaultMessageListenerContainer">
<property name="connectionFactory" ref="pooledConnectionFactory"/>
<property name="destination" ref="QUEUE_NAME"/>
<property name="messageListener" ref="productUpdateListener" />
<property name="sessionTransacted" value="true"/>
<property name="autoStartup" value="true"/>
</bean>
PooledConnectionFactory closes sessions/consumer connection after the consumer has processed messages.Hence the consumer count keeps reducing.
If you are using multithreaded environment its better to use CachingConnectionFactory. But this has some drawbacks as it caches the consumers so if a consumer hangs the client would keep pinging the stalled consumer without knowing that its connection is idle.
Exact solution for your problem, using JMS template is hard to find but you can manually create new connections by checking that if a connection has been closed by using connection.start() or connection.createConnection methods.
Related
Searching in the web I found a lot of example of how spring have to be configured (file.xml) for publish/subscriber mode for active MQ but I need how beans need to be configured for Websphere MQ
Anyone have been configured both product with success?
The following is bean definition of a Websphere MQ ConnectionFactory:
<bean id="wmqConnectionFactory" class="com.ibm.mq.jms.MQConnectionFactory">
<property name="queueManager">
<value>QUEUE_MANAGER</value>
</property>
<property name="hostName">
<value>127.0.0.1</value>
</property>
<property name="port">
<value>1414</value>
</property>
<property name="transportType">
<value>1</value>
</property>
<property name="channel">
<value>SYSTEM.DEF.SVRCONN</value>
</property>
</bean>
you may need to provide a username and password, and you can accomplish that by wrapping it in a UserCredentialsConnectionFactoryAdapter:
<bean id="userCredentialsAdapter" class="org.springframework.jms.connection.UserCredentialsConnectionFactoryAdapter">
<property name="targetConnectionFactory" ref="wmqConnectionFactory" />
<property name="username" value=" " />
<property name="password" value=" " />
</bean>
It is also a good idea to wrap it in a CachingConnectionFactory, so that sessions, consumers and producers are cached, this wrapper also switches the auto-recovery to true:
<bean id="cachingConnectionFactory" class="org.springframework.jms.connection.CachingConnectionFactory">
<property name="targetConnectionFactory" ref="userCredentialsAdapter"/>
<property name="cacheConsumers" value="true" />
<property name="cacheProducers" value="true" />
<property name="sessionCacheSize" value="10" />
</bean>
with this configuration, you can use the cachingConnectionFactory with whatever bean you need.
We are using Spring Integration in our project and we have a requirement where If IBM MQ goes down then we will have to auto connect to IBM MQ when it is up. We have done this implementation using recoveryInterval option of org.springframework.jms.listener.DefaultMessageListenerContainer class. We have given recovery interval as 6 seconds so every 6 seconds system try to recover the MQ connection but now we have a requirement where we will have to do the autorecover twice only and after that if still MQ is down then stop the inbound adapter.
Is there any way in Spring Integration to mention the auto recovery retry count so that system will try to recover only for that retry count?
Below is my existing configuration.
<bean id="inQ" class="com.ibm.mq.jms.MQQueue">
<constructor-arg value="${mq.inbound.queue}" />
</bean>
<int:channel id="inbound" />
<int-jms:message-driven-channel-adapter
id="jmsIn" channel="inbound" container="messageListenerContainer"
acknowledge="transacted" auto-startup="false">
</int-jms:message-driven-channel-adapter>
<int:service-activator id="mainService"
input-channel="inbound" ref="messageListener" method="onMessage">
<int:request-handler-advice-chain>
<ref bean="retryWithBackoffAdviceSession" />
</int:request-handler-advice-chain>
</int:service-activator>
<bean id="messageListenerContainer"
class="org.springframework.jms.listener.DefaultMessageListenerContainer">
<property name="connectionFactory" ref="mqConnectionFactory" />
<property name="destination" ref="inQ" />
<property name="sessionTransacted" value="true" />
<property name="maxConcurrentConsumers" value="${maxConcurrentConsumers}" />
<property name="concurrentConsumers" value="${concurrentConsumers}" />
<property name="receiveTimeout" value="${receiveTimeout}" />
<property name="recoveryInterval" value="60000" />
<property name="autoStartup" value="${autoStartup}" />
</bean>
Thanks
Sach
As an alternative to the recoveryInterval, you can now specify a Backoff instead (see the docs).
It doesn't provide a mechanism to stop the container but an appropriate backoff can effectively do what you want.
You would then need to programmatically stop/start to kick it off again.
We are using SpringFramework 3.2.12.RELEASE, Hibernate 4.1.6.FINAL against Oracle 11g.
We have a service managed by JPATransactionManager. The service accepts an id as primary key and creates a new entity with the id, saves it, makes an external call, then updates the entity. Under multi-thread testing with the same id, I am expecting all other threads are blocking after the first thread has done the insertion and making external call, however it is not the case. The service is annotated with #Transactional(propagation=Propagation.REQUIRES_NEW, readOnly=false), and the two save calls are made via JpaRepository. Below is my spring configuration.
<bean id="jpaVendorAdapter"
class="org.springframework.orm.jpa.vendor.HibernateJpaVendorAdapter">
<property name="showSql" value="true"></property>
<property name="generateDdl" value="false"></property>
<property name="database" value="ORACLE"></property>
</bean>
<bean id="abstractEntityManagerFactory"
class="org.springframework.orm.jpa.LocalContainerEntityManagerFactoryBean"
abstract="true">
<property name="jpaVendorAdapter" ref="jpaVendorAdapter" />
</bean>
<bean id="myEntityManagerFactory" parent="abstractEntityManagerFactory">
<property name="persistenceUnitName" value="my" />
<property name="dataSource" ref="myDataSource" />
<property name="PackagesToScan" value="test.schema" />
</bean>
<bean id="transactionManager" class="org.springframework.orm.jpa.JpaTransactionManager">
<property name="entityManagerFactory" ref="myEntityManagerFactory" />
</bean>
<tx:annotation-driven transaction-manager="transactionManager" />
<jpa:repositories base-package="test.dao"
entity-manager-factory-ref="myEntityManagerFactory"></jpa:repositories>
Appreciate comments and help!
I am trying to create an SSL connection to activemq (using Spring 3.1, ActiveMQ 5.5, and Camel 10.0). I'm getting the dreaded SSL handshake exception. I can connect with an openssl s_client using the certificate in the jks. Thus, I'm trying to figure out if there is a problem with my keystore (which seems to work in other situations - e.g. with tomcat) or a problem with my XML configuration. Does anyone have a good example, thoughts about other ways to test, or see what I'm doing wrong?
Please note, my config passes validation (in case you see a typo).
Thanks
-J
<bean id="myJmsRedeliverPolicy" class="org.apache.activemq.RedeliverPolicy">
<property name="maximumRedeliveries" value="500"/>
</bean>
<bean id="jmsSecureConnectionFactory" class="org.apache.activemq.ActiveMQSslConnectionFactory">
<property name="brokerURL" value="ssl://test.com:8100"/>
<property name="redeliverPolicy" ref="myJmsRedeliverPolicy"/>
<property name="keyStore" value="/usr/lib/mykeystore.jks"/>
<property name="keyStorePassword" value="mypass"/>
<property name="trustStore" value="/usr/lib/mycacerts"/>
<property name="trustStorePassword" value="changeit"/>
</bean>
<bean id="pooledSecureConnectionFactory" class="org.apache.activemq.pool.PooledConnectionFactory">
<property name="maxConnections" value="10"/>
<property name="connectionFactory" ref="jmsSecureConnectionFactory"/>
</bean>
<bean id="txSecureManager" class="org.springframework.jms.connection.JmsTransactionManager">
<property name="connectionFactory" ref="pooledSecureConnectionFactory" />
</bean>
<bean id="jmsSecureConfig" class="org.apache.camel.component.jms.JmsConfiguration">
<property name="connectionFactory" ref="pooledSecureConnectionFactory" />
<property name="testConnectionOnStartup" value="true"/>
<property name="transacted" value="true"/>
<property name="transactionManager" ref="txSecureManager"/>
</bean>
<bean id="activemqs" class="org.apache.activemq.camel.component.ActiveMQComponent">
<property name="configuration" ref="jmsSecureConfig" />
</bean>
Start your application with:
-Djavax.net.debug=ssl
to get further with your troubleshooting. Usually, that command provides printout that says pretty much exactly what's wrong.
My DAO integration tests are failing because entities created during the tests are still in the database at the start of the next test. The exact same behavior is seen from both MySQL 5 and H2.
The test classes are annotated with:
#Transactional
#RunWith(SpringJUnit4ClassRunner.class)
#ContextConfiguration( { "/testPersist-applicationContext.xml" })
The transaction bean configuration in the test application context is as follows:
<tx:annotation-driven />
<bean id="transactionManager"
class="org.springframework.transaction.jta.JtaTransactionManager">
<property name="transactionManager" ref="atomikosTransactionManager" />
<property name="userTransaction" ref="atomikosUserTransaction" />
</bean>
<bean id="atomikosTransactionManager" class="com.atomikos.icatch.jta.UserTransactionManager"
init-method="init" destroy-method="close">
<property name="forceShutdown" value="false" />
</bean>
<bean id="atomikosUserTransaction" class="com.atomikos.icatch.jta.UserTransactionImp">
<property name="transactionTimeout" value="300" />
</bean>
The entity manager is configured as follows:
<bean id="myappTestLocalEmf"
class="org.springframework.orm.jpa.LocalContainerEntityManagerFactoryBean">
<property name="persistenceUnitName" value="myapp" />
<property name="persistenceUnitPostProcessors">
<bean class="com.myapp.core.persist.util.JtaPersistenceUnitPostProcessor">
<property name="jtaDataSource" ref="myappPersistTestJdbcDataSource" />
</bean>
</property>
<property name="jpaVendorAdapter">
<bean class="org.springframework.orm.jpa.vendor.HibernateJpaVendorAdapter">
<property name="showSql" value="false" />
<property name="database" value="$DS{hibernate.database}" />
<property name="databasePlatform" value="$DS{hibernate.dialect}" />
</bean>
</property>
<property name="jpaProperties">
<props>
<prop key="hibernate.transaction.manager_lookup_class">com.atomikos.icatch.jta.hibernate3.TransactionManagerLookup
</prop>
<prop key="hibernate.format_sql">true"</prop>
<prop key="hibernate.use_sql_comments">true</prop>
</props>
</property>
</bean>
<context:annotation-config />
Everything in the log files seem to be fine...I can see messages from Spring about rollback and from Atomikos about rollback as well. But frankly, the logs are so huge and so complex, I could easily be missing something...
Yet the inserted test data remains! Any clues?
It turned out that my C3P0 JDBC data source was not XA aware and was therefore not participating in the transaction. Why I did not get an error, nor a warning in the log file, I do not know. Nonetheless, the reason you cannot use an XA aware data source is explained very nicely here. Note that the data source does not need to be XA capable...just XA aware.
Replacing the C3P0 data source with the following one solved the problem.
<bean id="myappJTANonXADataSource" class="com.atomikos.jdbc.nonxa.AtomikosNonXADataSourceBean">
<property name="uniqueResourceName" value="myappDatabase" />
<property name="driverClassName" value="$DS{hibernate.connection.driver_class}" />
<property name="url" value="$DS{hibernate.connection.url}" />
<property name="user" value="$DS{hibernate.connection.username}" />
<property name="password" value="$DS{hibernate.connection.password}" />
<property name="maxPoolSize" value="20" />
<property name="reapTimeout" value="300" />
</bean>
I think you will need to go through the logs in details. It could be that the rollbacks you are seeing are working, except that something else has executed a commit first. I also cannot see anything in your code which indicates automated rollback. And that it should occur at the end of each test. If you are depending on a timeout based rollback it could be that the second test is running before the timeout occurs, therefore it sees the data before it is rolled back.
Many options here there is :-)