在ActiveMQ队列上侦听的所有JMS使用者的油门消耗率 [英] Throttle consumption rate of all JMS consumers listening on an ActiveMQ queue

问题描述

使用骆驼节流阀组件，您可以定义每个时间段内已使用消息的速率的上限，例如：

Using Camel Throttler component you can define an upper limit on the rate of consumed messages per time period, e.g.:

from("activemq:queueA").throttle(10).to("direct-vm:bla")

我认为不理想的是，骆驼节流器在节流器阻塞的情况下将交换保持在内存中。因此，如果您有100个队列使用方，并且由于任何原因 direct-vm：bla运行缓慢（例如，调用外部慢速服务），则您的内存中最多可能有100个交换！

What I see as non ideal is that Camel Throttler keeps the exchanges in-memory while they are blocked by the Throttler. So, if you have 100 queue consumers and for any reason the "direct-vm:bla" is slow (e.g. invoking external slow service), you may have in-memory up to 100 exchanges!

我想知道是否有ActiveMQ功能将限制要求强加在ActiveMQ级别上。甚至在每个队列级别。类似于HornetQ的Consumer-max-rate。

I was wondering if there is an ActiveMQ feature to impose the throttling requirement on the ActiveMQ level. Maybe even on a per queue level. Something similar to consumer-max-rate of HornetQ.

推荐答案

配置消费者使用事务（不必XA事务。如果只有一个其他端点，请使用本地JMS事务（请参阅：LLR事务方法）。如果骆驼路线失败，则消息回滚到代理。

Configure the consumer to use a transaction (doesn't have to be an XA transaction. If you only have 1 other endpoint use a local JMS transaction. see: LLR transaction approach). If the Camel route fails, the messages roll back to the broker.

这篇关于在ActiveMQ队列上侦听的所有JMS使用者的油门消耗率的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！