<html><head><meta http-equiv="Content-Type" content="text/html; charset=UTF-8"></head><body><div>Hello Simon, </div><div> Oops. I meant to say I removed acks. </div><div><br></div><div>Shri</div><div><br></div><div><br></div> <br>Simon MacMullen <simon@rabbitmq.com> wrote:<br>Are you saying you turned *off* auto-ack - i.e. you turned <br>acknowledgements on? If so, that would account for greater network and <br>CPU use.<br><br>Cheers, Simon<br><br>On 04/12/2013 23:11, shridharan muthu wrote:<br>> Hello there,<br>> As a performance optimization on our rabbit mq cluster (of 2<br>> nodes), I removed following from a high volume non-durable queue<br>><br>> 1. AMQP_MANDATORY flag during publish and<br>> 2. AMQP_AUTOACK during consume.<br>><br>> As a result the number of messages on that queue went down from 150 to<br>> 15 which is fantastic and expected. But this also increased overall cpu<br>> usage from 22% to 32% with an increased network traffic from 30k<br>> packets/sec to 40k packets/sec. The ingress & egress rate stayed the<br>> same before and after the change i.e. around 300 msgs/sec.<br>><br>> So, my question is why would network traffic increase when I make all<br>> calls asynchronous when the ingress & egress rate is same ?<br>><br>> More info about the cluster:<br>> shri@mq1:~$ sudo rabbitmqctl status<br>> Status of node rabbit@mq1 ...<br>> [{pid,11689},<br>> {running_applications,<br>> [{rabbitmq_management,"RabbitMQ Management Console","3.0.2"},<br>> {rabbitmq_management_agent,"RabbitMQ Management Agent","3.0.2"},<br>> {rabbit,"RabbitMQ","3.0.2"},<br>> {os_mon,"CPO CXC 138 46","2.2.4"},<br>> {rabbitmq_mochiweb,"RabbitMQ Mochiweb Embedding","3.0.2"},<br>> {webmachine,"webmachine","1.9.1-rmq3.0.2-git52e62bc"},<br>> {mochiweb,"MochiMedia Web Server","2.3.1-rmq3.0.2-gitd541e9a"},<br>> {mnesia,"MNESIA CXC 138 12","4.4.12"},<br>> {amqp_client,"RabbitMQ AMQP Client","3.0.2"},<br>> {xmerl,"XML parser","1.2.3"},<br>> {inets,"INETS CXC 138 49","5.2"},<br>> {sasl,"SASL CXC 138 11","2.1.8"},<br>> {stdlib,"ERTS CXC 138 10","1.16.4"},<br>> {kernel,"ERTS CXC 138 10","2.13.4"}]},<br>> {os,{unix,linux}},<br>> {erlang_version,<br>> "Erlang R13B03 (erts-5.7.4) [source] [64-bit] [smp:12:12] [rq:12]<br>> [async-threads:30] [hipe] [kernel-poll:true]\n"},<br>> {memory,<br>> [{total,469535752},<br>> {connection_procs,17995944},<br>> {queue_procs,11675696},<br>> {plugins,230048},<br>> {other_proc,126014264},<br>> {mnesia,18446743803509563784},<br>> {mgmt_db,12530864},<br>> {msg_index,1934440},<br>> {other_ets,-18446743803280937440},<br>> {binary,5524296},<br>> {code,18395964},<br>> {atom,3440321},<br>> {other_system,43167571}]},<br>> {vm_memory_high_watermark,0.4},<br>> {vm_memory_limit,5047128883},<br>> {disk_free_limit,1000000000},<br>> {disk_free,482189381632},<br>> {file_descriptors,<br>> [{total_limit,65436},<br>> {total_used,309},<br>> {sockets_limit,58890},<br>> {sockets_used,298}]},<br>> {processes,[{limit,1048576},{used,2992}]},<br>> {run_queue,7},<br>> {uptime,23771203}]<br>> ...done.<br>><br>><br>> Thank you,<br>> Shri<br>><br>><br>> _______________________________________________<br>> rabbitmq-discuss mailing list<br>> rabbitmq-discuss@lists.rabbitmq.com<br>> https://lists.rabbitmq.com/cgi-bin/mailman/listinfo/rabbitmq-discuss<br>><br><br></body>