# Ideal CPU spec configuration for a 2 node and a 4 node cluster

**URL:** <https://discuss.elastic.co/t/ideal-cpu-spec-configuration-for-a-2-node-and-a-4-node-cluster/100963>\
**Category:** Elasticsearch\
**Created:** [September 19, 2017, 6:22am UTC](https://discuss.elastic.co/t/ideal-cpu-spec-configuration-for-a-2-node-and-a-4-node-cluster/100963 "2017-09-19T06:22:33Z")\
**Posts on this page:** 5\
**Page:** 1

<div class="post-metadata">

**Author:** ![iam.Carrot](https://avatars.discourse-cdn.com/v4/letter/i/a9adbd/32.png) [@iam.Carrot](https://discuss.elastic.co/u/iam.Carrot)\
**Post date:** [September 19, 2017, 6:22am UTC](https://discuss.elastic.co/t/ideal-cpu-spec-configuration-for-a-2-node-and-a-4-node-cluster/100963/1 "2017-09-19T06:22:33Z")

</div>

Hi, I am working on elasticsearch and I would want to scale up to an ideal setup for a 2 node or a 4 node cluster. Can someone please inform me about what should be the idea CPU configuration for:

## 2 Nodes:

1. Both nodes would work as all roles (serve request and hold data).
2. The nodes would have 5 shards and 1 replica.
3. The nodes would be hosted as VM's on AWS

## 4 nodes:

In a dedicated hierarchical architecture:  
`1 Master` node, `1 Coordinate only` node and `2 data` nodes.

* * *

## What I want to know?

1. What should be a ideal CPU specs (RAM? Cores? Disk?) for each of my nodes for each of the above mentioned configuration.
2. How many requests per minute/second would my cluster be able to handle. For example what kind of an infra am I looking at if I have a throughput of `500`, `1000` and `5000` requests in a second

* * *

I am using elasticsearch only for searching purposes. But it's quite an extensive search through

---

<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:** [September 19, 2017, 7:38am UTC](https://discuss.elastic.co/t/ideal-cpu-spec-configuration-for-a-2-node-and-a-4-node-cluster/100963/2 "2017-09-19T07:38:38Z")

</div>

> [@iam.Carrot](#):
>
> 1 Master node, 1 Coordinate only node and 2 data nodes.

That's not ideal, see [Important Configuration Changes | Elasticsearch: The Definitive Guide [2.x] | Elastic](https://www.elastic.co/guide/en/elasticsearch/guide/2.x/important-configuration-changes.html#_minimum_master_nodes)

1. Depends on what use case you are talking about
2. See 1. Also, testing is the only thing that can definitively tell you this

---

<div class="post-metadata">

**Author:** ![iam.Carrot](https://avatars.discourse-cdn.com/v4/letter/i/a9adbd/32.png) [@iam.Carrot](https://discuss.elastic.co/u/iam.Carrot)\
**Post date:** [September 19, 2017, 8:09am UTC](https://discuss.elastic.co/t/ideal-cpu-spec-configuration-for-a-2-node-and-a-4-node-cluster/100963/3 "2017-09-19T08:09:57Z")

</div>

@warkolm I do understand about the minimum 3 master eligible nodes but as of this moment in scaling up i can only scale up to the two defined set ups. Would you be able to share the CPU specs for it? Like how much ram, cores should I go for to be able to set up a stable and fast searching engine.

I have roughly 10 million documents stored in 5 shards and 1 replica of each.

I am using elasticsearch for an autosuggest set up that means roughly 1000 requests would come in a second and the response times have to lower than 100ms preferably below 50ms. Currently on a single node I get a response time of around 86ms with 100 requests per second.

---

<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:** [September 19, 2017, 8:11am UTC](https://discuss.elastic.co/t/ideal-cpu-spec-configuration-for-a-2-node-and-a-4-node-cluster/100963/4 "2017-09-19T08:11:35Z")

</div>

> [@iam.Carrot](#):
>
> I do understand about the minimum 3 master eligible nodes but as of this moment in scaling up i can only scale up to the two defined set ups

Make all 3 nodes master eligible then 🙂

> [@iam.Carrot](#):
>
> Would you be able to share the CPU specs for it? Like how much ram, cores should I go for to be able to set up a stable and fast searching engine.

Like I said, you need to test based on the SLAs you have defined.

Start with 4 CPUs, go from there.

---

<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:** [October 17, 2017, 8:11am UTC](https://discuss.elastic.co/t/ideal-cpu-spec-configuration-for-a-2-node-and-a-4-node-cluster/100963/5 "2017-10-17T08:11:43Z")

</div>

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