骆驼政策

Question

我有一条小路线，我想使用自定义的重新传递策略来重复向endpoint发送消息，但这种行为非常奇怪。看起来，重新交付政策只是在重复一个错误。我试图将所有交换发送到路由的开头，但策略不起作用，因为每次都在创建：

        <route id="sampleRoute">
            <from uri="direct:anotheRoute" />
            <to uri="nmr:kaboom" />
            <choice>
                <when>
                    <xpath>//result = 'true'</xpath>
                    <to uri="direct:anotherAnotherRoute" />
                </when>
                <otherwise>
                    <throwException ref="redeliveryException" />
                </otherwise>
            </choice>
            <onException>
                <exception>java.net.SocketException</exception>
                <exception>java.net.ConnectException</exception>
                <exception>my.custom.error.RedeliveryException</exception>
                <redeliveryPolicy ref="customRedeliveryPolicy" />
                <to uri="direct:anotheRoute" />
            </onException>
        </route>

我做错了什么？当错误发生时，我想以间隔重复我的请求。我的骆驼版本是2.6

日志：

16:29:43,780 | DEBUG | cc5bd91-thread-1 | DefaultListableBeanFactory       |  | 59 - org.springframework.beans - 3.0.5.RELEASE | Returning cached instance of singleton bean 'RedeliveryPolicy'
16:29:43,780 | DEBUG | cc5bd91-thread-1 | DefaultErrorHandler              |  | 68 - org.apache.camel.camel-core - 2.6.0.fuse-03-01 | Failed delivery for exchangeId: ID-1NVC-52527-1436958153799-314-4. On delivery attempt: 0 caught: my.custom.error.RedeliveryException
16:29:43,780 | DEBUG | cc5bd91-thread-1 | RedeliveryPolicy                 |  | 68 - org.apache.camel.camel-core - 2.6.0.fuse-03-01 | Sleeping for: 60 millis until attempting redelivery
16:29:43,840 | DEBUG | cc5bd91-thread-1 | DefaultListableBeanFactory       |  | 59 - org.springframework.beans - 3.0.5.RELEASE | Returning cached instance of singleton bean 'RedeliveryPolicy'
16:29:43,840 | DEBUG | cc5bd91-thread-1 | DefaultErrorHandler              |  | 68 - org.apache.camel.camel-core - 2.6.0.fuse-03-01 | Failed delivery for exchangeId: ID-1NVC-52527-1436958153799-314-4. On delivery attempt: 1 caught: my.custom.error.RedeliveryException
16:29:43,840 | DEBUG | cc5bd91-thread-1 | RedeliveryPolicy                 |  | 68 - org.apache.camel.camel-core - 2.6.0.fuse-03-01 | Sleeping for: 60 millis until attempting redelivery
16:29:43,900 | DEBUG | cc5bd91-thread-1 | DefaultListableBeanFactory       |  | 59 - org.springframework.beans - 3.0.5.RELEASE | Returning cached instance of singleton bean 'RedeliveryPolicy'
16:29:43,900 | DEBUG | cc5bd91-thread-1 | DefaultErrorHandler              |  | 68 - org.apache.camel.camel-core - 2.6.0.fuse-03-01 | Failed delivery for exchangeId: ID-1NVC-52527-1436958153799-314-4. On delivery attempt: 2 caught: my.custom.error.RedeliveryException
16:29:43,900 | DEBUG | cc5bd91-thread-1 | RedeliveryPolicy                 |  | 68 - org.apache.camel.camel-core - 2.6.0.fuse-03-01 | Sleeping for: 60 millis until attempting redelivery
16:29:43,960 | DEBUG | cc5bd91-thread-1 | DefaultListableBeanFactory       |  | 59 - org.springframework.beans - 3.0.5.RELEASE | Returning cached instance of singleton bean 'RedeliveryPolicy'
16:29:43,960 | DEBUG | cc5bd91-thread-1 | DefaultErrorHandler              |  | 68 - org.apache.camel.camel-core - 2.6.0.fuse-03-01 | Failed delivery for exchangeId: ID-1NVC-52527-1436958153799-314-4. On delivery attempt: 3 caught: my.custom.error.RedeliveryException
16:29:43,960 | DEBUG | cc5bd91-thread-1 | RedeliveryPolicy                 |  | 68 - org.apache.camel.camel-core - 2.6.0.fuse-03-01 | Sleeping for: 60 millis until attempting redelivery
16:29:44,020 | DEBUG | cc5bd91-thread-1 | DefaultListableBeanFactory       |  | 59 - org.springframework.beans - 3.0.5.RELEASE | Returning cached instance of singleton bean 'RedeliveryPolicy'
16:29:44,020 | DEBUG | cc5bd91-thread-1 | DefaultErrorHandler              |  | 68 - org.apache.camel.camel-core - 2.6.0.fuse-03-01 | Failed delivery for exchangeId: ID-1NVC-52527-1436958153799-314-4. On delivery attempt: 4 caught: my.custom.error.RedeliveryException
16:29:44,020 | DEBUG | cc5bd91-thread-1 | RedeliveryPolicy                 |  | 68 - org.apache.camel.camel-core - 2.6.0.fuse-03-01 | Sleeping for: 60 millis until attempting redelivery
16:29:44,080 | DEBUG | cc5bd91-thread-1 | DefaultListableBeanFactory       |  | 59 - org.springframework.beans - 3.0.5.RELEASE | Returning cached instance of singleton bean 'RedeliveryPolicy'
16:29:44,080 | DEBUG | cc5bd91-thread-1 | DefaultErrorHandler              |  | 68 - org.apache.camel.camel-core - 2.6.0.fuse-03-01 | Failed delivery for exchangeId: ID-1NVC-52527-1436958153799-314-4. On delivery attempt: 5 caught: my.custom.error.RedeliveryException
16:29:44,081 | DEBUG | cc5bd91-thread-1 | NMR                              |  | 81 - org.apache.servicemix.nmr.core - 1.4.0.fuse-03-01 | Channel org.apache.servicemix.nmr.core.ChannelImpl@7dc124e5 dispatching exchange: [

Answer 1

谢谢你的解释。我将发布正确答案：

        <route id="sampleRoute">
            <from uri="direct:anotheRoute" />
            <to uri="direct:kaboom" />
            <onException>
                <exception>java.net.SocketException</exception>
                <exception>java.net.ConnectException</exception>
                <exception>my.custom.error.RedeliveryException</exception>
                <redeliveryPolicy ref="customRedeliveryPolicy" />
            </onException>
        </route>
        <route errorHandlerRef="noErrorHandler">
            <from uri="direct:kaboom />
            <to uri="nmr:kaboom/>
            <choice>
                <when>
                    <xpath>//result = 'true'</xpath>
                    <to uri="direct:anotherAnotherRoute" />
                </when>
                <otherwise>
                    <throwException ref="redeliveryException" />
                </otherwise>
            </choice>
        </route>

豆子里的声明：

<bean id="noErrorHandler" class="org.apache.camel.builder.NoErrorHandlerBuilder"/>

Answer 2

这并不奇怪，Camel的错误处理程序和重新交付是从失败的步骤开始执行的，而不是从路线的开始。

你可以不用再次抛出错误调用direct:anotherRoute，唯一的问题是如果你这么做太频繁，你的调用堆栈可能会变得太深。

你可以做的是把那个nmr： kaboom放到一个单独的路由中，并将其配置为具有noErrorHandler，然后从第一个路由开始，用Direct：调用它，然后错误处理程序可以从头重新传递整个路由。