# Multiple Logstash outputs vs pipeline-to-pipeline communcations

**URL:** <https://discuss.elastic.co/t/multiple-logstash-outputs-vs-pipeline-to-pipeline-communcations/273546>\
**Category:** Logstash\
**Created:** [May 20, 2021, 1:24pm UTC](https://discuss.elastic.co/t/multiple-logstash-outputs-vs-pipeline-to-pipeline-communcations/273546 "2021-05-20T13:24:25Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Kosodrom](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/kosodrom/32/99278_2.png) [@Kosodrom](https://discuss.elastic.co/u/Kosodrom)\
**Post date:** [May 20, 2021, 1:24pm UTC](https://discuss.elastic.co/t/multiple-logstash-outputs-vs-pipeline-to-pipeline-communcations/273546/1 "2021-05-20T13:24:25Z")

</div>

Hi folks,

I have following use case:

I want to create one index for each kubernetes namespace (and its own lifecycle policy) and I want to use the ILM. As far as I understand I have two approaches:

1. I have one pipeline config file, lets say `kubernetes-logs.conf` in which I have an output consisting of if else statements and the respective index settings for the namespace logs such as rollover\_alias, ilm\_pattern.
2. I have different pipeline config files, one per namespace. And I use pipeline-to-pipeline communication to redirect the traffic to different pipelines inside the `pipelines.yml`

Do you have any hints which approach is better in terms of ressource consumption?

Thanks

---

<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:** [June 17, 2021, 1:24pm UTC](https://discuss.elastic.co/t/multiple-logstash-outputs-vs-pipeline-to-pipeline-communcations/273546/2 "2021-06-17T13:24:47Z")

</div>

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.