# Logstash rabbitmq output performance

**URL:** <https://discuss.elastic.co/t/logstash-rabbitmq-output-performance/59152>\
**Category:** Logstash\
**Created:** [August 29, 2016, 6:44am UTC](https://discuss.elastic.co/t/logstash-rabbitmq-output-performance/59152 "2016-08-29T06:44:38Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![lkosz](https://avatars.discourse-cdn.com/v4/letter/l/8491ac/32.png) [@lkosz](https://discuss.elastic.co/u/lkosz)\
**Post date:** [August 29, 2016, 6:44am UTC](https://discuss.elastic.co/t/logstash-rabbitmq-output-performance/59152/1 "2016-08-29T06:44:38Z")

</div>

Hi,

I have ELK in following scheme:  
`[log-courier on hosts] ~> [logstash-cache (15 instances)] ~> [haproxy] ~> [rabbitmq logstash-cache queue] ~> [haproxy] ~> [logstash-indexer (5 instances)] ~> [haproxy] ~> [rabbitmq logstash-writer queue] ~> [haproxy] ~> [logstash-writer (5 instances)]`

Here is my logstash-cache config:  
[http://pastebin.com/HdE4iWHw](http://pastebin.com/HdE4iWHw)

Everything is ok except one problem - can receive 4000 doc/s tops. When I set output to file only, single logstash-cache instance can store \>35k doc/s. I know that rabbit protocol have his cost, but 4000 events/s from 15 instances seems be too few. How to increase performance?

---

<div class="post-metadata">

**Author:** ![system](https://us1.discourse-cdn.com/elastic/original/3X/1/a/1ac57faf039f6b580b3f104ef42a2a89e41014de.png) [@system](https://discuss.elastic.co/u/system)\
**Post date:** [July 6, 2017, 4:41am UTC](https://discuss.elastic.co/t/logstash-rabbitmq-output-performance/59152/2 "2017-07-06T04:41:13Z")

</div>