I developed program to publish messages and then consume it at 10 ms interval for 30 s duration. When it runs for 5 s, it will disconnect and reconnect after 5 s but still produce messages. Apache Qpid need at least 4 s to consume all the messages but RabbitMQ only need 2 s. I guessed there is maximum bandwidth to publish messages. What caused this? Is amqp 0.9.1 lighter then amqp 1.0? Or there are any reasons?
Asked
Active
Viewed 104 times
0
-
Could you be a bit more specific and add how you've measured those times? – JiriS Nov 04 '17 at 22:25
-
I have 2 virtual os, 1 as a client (to publish and subscribe) and another one as server (broker for amqp). client will generate message with interval 10 ms then, after around 5 seconds client will connect to server. There are around 250 - 300 messages (1024 byte each). Qpid need almost 4 seconds to consume and Rabbitmq only around 2 seconds. What caused this difference? Is there any bandwidth limitation between amqp 0.9.1 and amqp 1.0? or it just config problem in Qpid or Rabbitmq? thanks – Faiqul Azmi Nov 04 '17 at 22:52
-
The messaging delivery guarantee offered by the messaging provider to your application has a huge bearing on overall performance of the system. It influences the messaging packets that need to flow to and fro over the network and how the messaging provider uses storage. If you have not done so already, you need to first decide what delivery guarantee your application requires, then make sure you configure Rabbit or Qpid to provide the guarantee you require. This usually mean configuration on both the client and server sides. This will allow you to conduct a fair test. – k-wall Nov 06 '17 at 08:50