Hi,
I have a rabbitmq cluster with three nodes with a HAProxy
above them. The heartbeat interval is 30 sec for the nodes. When I look
into the logs(debug enabled), I'm seeing frequent closing of AMQP
connection.When I browsed through internet,
I found that "
connection_closed_abruptly" is due to client closing connection during TCP connection. Is
"connection_closed_with_no_data_received" due to heartbeat? The logs are flooded with the below mentioned pattern. (There are around 700 active connections).
I believe, as a side effect, the publishing of message is exceeding 500 ms(1.3 sec max as of now, publishing frequency ~2500 msg/sec). I really want this publish time to belesser than 500ms. Is there any design which would ensures this. I'm planning to increase the heartbeat and see how things pan out. Any input on this would be greatly appreciated. Sorry, if this is a dumb question. See the RabbitMQ logs below.
=INFO REPORT==== 26-May-2015::12:59:50 ===
accepting AMQP connection <0.26633.59> (10.81.198.175:42605 -> 10.11.166.194:9999)
=WARNING REPORT==== 26-May-2015::12:59:50 ===
closing AMQP connection <0.26633.59> (10.81.198.175:42605 -> 10.11.166.194:9999):
connection_closed_abruptly
=INFO REPORT==== 26-May-2015::12:59:50 ===
accepting AMQP connection <0.26644.59> (10.81.198.175:42608 -> 10.11.166.194:9999)
=INFO REPORT==== 26-May-2015::12:59:50 ===
closing AMQP connection <0.26644.59> (10.81.198.175:42608 -> 10.11.166.194:9999):
connection_closed_with_no_data_received
=INFO REPORT==== 26-May-2015::12:59:51 ===
accepting AMQP connection <0.26649.59> (10.81.198.175:42612 -> 10.11.166.194:9999)
Thanks
Manoj