# Pointing to different Nodes

**URL:** <https://discuss.elastic.co/t/pointing-to-different-nodes/251198>\
**Category:** Elasticsearch\
**Created:** [October 7, 2020, 12:49am UTC](https://discuss.elastic.co/t/pointing-to-different-nodes/251198 "2020-10-07T00:49:23Z")\
**Posts on this page:** 5\
**Page:** 1

<div class="post-metadata">

**Author:** ![Jtomage](https://avatars.discourse-cdn.com/v4/letter/j/e5b9ba/32.png) [@Jtomage](https://discuss.elastic.co/u/Jtomage)\
**Post date:** [October 7, 2020, 12:49am UTC](https://discuss.elastic.co/t/pointing-to-different-nodes/251198/1 "2020-10-07T00:49:23Z")

</div>

Hi all,

New to elasticsearch I am currently for clarification on pointing to nodes with different roles. The current set up is a 3 Master-data node cluster and looking to possibly scale out. Currently there are 2 applications,

1. that continuously writes to elasticsearch
2. a web application for users to search.

According to this post [correct-usage-of-coordinating-role-node](https://discuss.elastic.co/t/correct-usage-of-coordinating-role-node/109988/13) I can the application point directly to the coordination node or data node via the IP address / url of elasticsearch node.

Meaning for application

1. the writing application, I create connection pool of urls / ip address of **only** the data nodes and send the data their.
2. the web app, I can add a coordination node to cluster and use **only** the ip address / url of the coordination node

Then if an ingest node is added to the cluster, do I add that nodes url / ip address to the connection pool of the data nodes?

Thank you

---

<div class="post-metadata">

**Author:** ![rschirin](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/rschirin/32/45283_2.png) [@rschirin](https://discuss.elastic.co/u/rschirin)\
**Post date:** [October 7, 2020, 6:51am UTC](https://discuss.elastic.co/t/pointing-to-different-nodes/251198/2 "2020-10-07T06:51:40Z")

</div>

if your new node (with ingest role) is also a data node, you can add it. otherwise not.

---

<div class="post-metadata">

**Author:** ![Jtomage](https://avatars.discourse-cdn.com/v4/letter/j/e5b9ba/32.png) [@Jtomage](https://discuss.elastic.co/u/Jtomage)\
**Post date:** [October 7, 2020, 3:43pm UTC](https://discuss.elastic.co/t/pointing-to-different-nodes/251198/3 "2020-10-07T15:43:32Z")

</div>

Did I understand correctly about pointing the different applications to the different nodes?

Interesting, I do not need to point to an ingest only node from the application as elasticsearch will handle it.

---

<div class="post-metadata">

**Author:** ![rschirin](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/rschirin/32/45283_2.png) [@rschirin](https://discuss.elastic.co/u/rschirin)\
**Post date:** [October 8, 2020, 7:36am UTC](https://discuss.elastic.co/t/pointing-to-different-nodes/251198/4 "2020-10-08T07:36:17Z")

</div>

from my personal point of view, you should use coordinating-only role for both activities (search and bulk-indexing actions).

following doc, [here](https://www.elastic.co/guide/en/elasticsearch/reference/master/ingest.html), ingest node is used to preprocess docs. if you do not need the preprocessing phase, you can skip it.

---

<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:** [November 5, 2020, 7:36am UTC](https://discuss.elastic.co/t/pointing-to-different-nodes/251198/5 "2020-11-05T07:36:36Z")

</div>

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