# Ingest only nodes

**URL:** <https://discuss.elastic.co/t/ingest-only-nodes/235854>\
**Category:** Elasticsearch\
**Created:** [June 4, 2020, 10:17pm UTC](https://discuss.elastic.co/t/ingest-only-nodes/235854 "2020-06-04T22:17:41Z")\
**Posts on this page:** 5\
**Page:** 1

<div class="post-metadata">

**Author:** ![barracuda-charles](https://avatars.discourse-cdn.com/v4/letter/b/278dde/32.png) [@barracuda-charles](https://discuss.elastic.co/u/barracuda-charles)\
**Post date:** [June 4, 2020, 10:17pm UTC](https://discuss.elastic.co/t/ingest-only-nodes/235854/1 "2020-06-04T22:17:41Z")

</div>

I have an ingest-heavy workload (many more writes than reads). I am curious about the need / benefits of having ingest-only nodes vs having the ingest done on the data nodes.

I don't think I actually have a specific ingest pipeline, but my question is whether or not things like indexing / analyzing (e.g. stemming) happens on the ingest nodes or not. Is it beneficial for me to offload this processing to ingest-only nodes, or is that only suggested if you have a specific "ingest pipeline"?

Thanks in advance!

---

<div class="post-metadata">

**Author:** ![warkolm](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/warkolm/32/39224_2.png) [@warkolm](https://discuss.elastic.co/u/warkolm)\
**Post date:** [June 5, 2020, 12:06am UTC](https://discuss.elastic.co/t/ingest-only-nodes/235854/2 "2020-06-05T00:06:27Z")

</div>

Sounds like it yes, especially if you can customise the hardware to give it more CPU (and not as much disk).

---

<div class="post-metadata">

**Author:** ![Christian\_Dahlqvist](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/christian_dahlqvist/32/4617_2.png) [@Christian\_Dahlqvist](https://discuss.elastic.co/u/Christian_Dahlqvist)\
**Post date:** [June 5, 2020, 5:23am UTC](https://discuss.elastic.co/t/ingest-only-nodes/235854/3 "2020-06-05T05:23:25Z")

</div>

Dedicated ingest nodes are generally only required if you are using ingest pipelines. Without pipelines they will acy as coordinating nodes which can be beneficial but does not do a lot of work. indexing and analysis is all done on the data nodes.

---

<div class="post-metadata">

**Author:** ![barracuda-charles](https://avatars.discourse-cdn.com/v4/letter/b/278dde/32.png) [@barracuda-charles](https://discuss.elastic.co/u/barracuda-charles)\
**Post date:** [June 6, 2020, 4:28pm UTC](https://discuss.elastic.co/t/ingest-only-nodes/235854/4 "2020-06-06T16:28:47Z")

</div>

Great -- that's what I was thinking but needed verification. Much thanks for the info!

---

<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 4, 2020, 4:28pm UTC](https://discuss.elastic.co/t/ingest-only-nodes/235854/5 "2020-07-04T16:28:50Z")

</div>

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