# CPU Affinity and Numa

**URL:** <https://discuss.elastic.co/t/cpu-affinity-and-numa/114576>\
**Category:** Elasticsearch\
**Created:** [January 8, 2018, 5:24pm UTC](https://discuss.elastic.co/t/cpu-affinity-and-numa/114576 "2018-01-08T17:24:00Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![morphers82](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/morphers82/32/22904_2.png) [@morphers82](https://discuss.elastic.co/u/morphers82)\
**Post date:** [January 8, 2018, 5:24pm UTC](https://discuss.elastic.co/t/cpu-affinity-and-numa/114576/1 "2018-01-08T17:24:01Z")

</div>

I am hosting multiple nodes on single servers that include multiple numa nodes.

I would like to setup some kind of cpu/core affinity for each elasticsearch process and a better numa policy than the default.  
I am on ubuntu 16.04.  
I am hoping this will increase performance of indexing/search.

I can't find much documentation on this, especially specifically for elasticsearch.

I already set the number of processors in the yml file to the number of total available threads divided by the number of nodes on the server.

I think there has to be some setting i have to put into the jvm.options file for cpu affinity, just have no idea what it would be?

---

<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:** [February 5, 2018, 5:24pm UTC](https://discuss.elastic.co/t/cpu-affinity-and-numa/114576/2 "2018-02-05T17:24:01Z")

</div>

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