# Elasticsearch cluster production deployment

**URL:** <https://discuss.elastic.co/t/elasticsearch-cluster-production-deployment/108451>\
**Category:** Elasticsearch\
**Created:** [November 20, 2017, 8:11pm UTC](https://discuss.elastic.co/t/elasticsearch-cluster-production-deployment/108451 "2017-11-20T20:11:46Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![pri](https://avatars.discourse-cdn.com/v4/letter/p/22d042/32.png) [@pri](https://discuss.elastic.co/u/pri)\
**Post date:** [November 20, 2017, 8:11pm UTC](https://discuss.elastic.co/t/elasticsearch-cluster-production-deployment/108451/1 "2017-11-20T20:11:46Z")

</div>

Hi,  
Have a couple of questions regarding the configuration for ES production cluster.  
We have a production cluster with 2 data nodes, 3 master eligible nodes and 1 coordinating node. All are configured as dedicated nodes.

1. Have read that the recommended setting for replica shards is 2 for production. Since we have only 2 data nodes, does it still make sense to setting the replicas to 2? Since there can be only one replica that can actually be of any use if a data node were to go down.
2. If we decide to add one more data node to the cluster later on, will the existing shards and replicas automatically redistribute without down time? And also is there a time delay before the replicas or primary shards become available in the new data node?
3. Is it a normal practice to horizontally scale coordinating nodes?  
Thanks for you time and help.

---

<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:** [November 20, 2017, 8:17pm UTC](https://discuss.elastic.co/t/elasticsearch-cluster-production-deployment/108451/2 "2017-11-20T20:17:23Z")

</div>

1. Where did you read that? No, it doesn't make sense for your setup.
2. Yes. The delay is purely the time it takes to rebalance, but your data is still fully accessible.
3. Depends. If you have the infrastructure, sure.

---

<div class="post-metadata">

**Author:** ![pri](https://avatars.discourse-cdn.com/v4/letter/p/22d042/32.png) [@pri](https://discuss.elastic.co/u/pri)\
**Post date:** [November 21, 2017, 3:48am UTC](https://discuss.elastic.co/t/elasticsearch-cluster-production-deployment/108451/3 "2017-11-21T03:48:11Z")

</div>

Thank you for the answers. Both the blogs - the ones that said the recommended setting for replicas in production is 2 - are not from elastic. Thanks for clarifying.

---

<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:** [December 19, 2017, 3:48am UTC](https://discuss.elastic.co/t/elasticsearch-cluster-production-deployment/108451/4 "2017-12-19T03:48:28Z")

</div>

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