Each of the consumer routes is brought up concurrently on different threads. They will suck in messages as fast as they can, competing like kids for candy tossed on the middle of a floor. So while thread 1 is waiting 5 seconds, threads 2, 3, 4, 5, ... and so on aren't and grabbing the next message on the queue. When one of those is waiting, the others are probably not. The more concurrent consumers you put on the queue, the less effect your delay timer will have.
Out of curiosity why are you doing this? There are better ways of throttling if that is your goal. *Robert Simmons Jr. MSc. - Lead Java Architect @ EA* *Author of: Hardcore Java (2003) and Maintainable Java (2012)* *LinkedIn: **http://www.linkedin.com/pub/robert-simmons/40/852/a39 <http://www.linkedin.com/pub/robert-simmons/40/852/a39>* On Tue, Dec 10, 2013 at 10:58 PM, Dimitry Pletnikov <[email protected]>wrote: > Using: > - ServiceMix 4.3.5 > - ActiveMQ 5.7.0 > - Camel 2.10.7 > > I have a persistent queue in ActiveMQ, which I consume messages from > using activemq-component with concurrentConsumers=20 set. > > ActiveMQ connections are pooled: > > <bean id="activemqConnectionFactory" > class="org.apache.activemq.ActiveMQConnectionFactory" > depends-on="broker"> > > <property name="brokerURL" > value="vm://${activemq.brokername}?create=false&waitForStart=10000" > /> > <property name="userName" value="${activemq.user}"/> > <property name="password" value="${activemq.password}"/> > </bean> > > <bean id="pooledConnectionFactory" > class="org.apache.activemq.pool.PooledConnectionFactory"> > <property name="idleTimeout" value="0" /> > <property name="maxConnections" value="8" /> > <property name="connectionFactory" ref="activemqConnectionFactory" > /> > </bean> > > <bean id="resourceManager" > class="org.apache.activemq.pool.ActiveMQResourceManager" > init-method="recoverResource"> > <property name="transactionManager" ref="recoverableTxManager" /> > <property name="connectionFactory" ref="activemqConnectionFactory" > /> > <property name="resourceName" value="activemq.${name}" /> > </bean> > > <reference id="recoverableTxManager" > > interface="org.apache.geronimo.transaction.manager.RecoverableTransactionManager" > availability="mandatory" /> > > I have a route looking like: > > <from uri="activemq://item_updates?concurrentConsumers=20" /> > <delay><constant>5000</constant></delay> > <log message="item ${header.item_id}" /> > > I expect to see something like: > item 1 > item 2 > item 3 > item 4 > item 5 > <5 seconds pause> > item 6 > item 7 > ..etc > > Instead the output looks like: > item 1 > <5 seconds pause> > item 2 > <5 seconds pause> > ..etc > > My guesses are: > - all messages are being prefetched by the first concurrent consumer? > -or- > - it has something to do with pooled ActiveMQ connections? > -or- > - ??? > > Looking at examples and documentation it should be simple and easy. In > my case it's not. >
