# ES 5.1 Server Nodes - SSD vs CEPH

**URL:** <https://discuss.elastic.co/t/es-5-1-server-nodes-ssd-vs-ceph/70008>\
**Category:** Elasticsearch\
**Created:** [December 26, 2016, 1:43pm UTC](https://discuss.elastic.co/t/es-5-1-server-nodes-ssd-vs-ceph/70008 "2016-12-26T13:43:38Z")\
**Posts on this page:** 5\
**Page:** 1

<div class="post-metadata">

**Author:** ![evert](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/evert/32/8143_2.png) [@evert](https://discuss.elastic.co/u/evert)\
**Post date:** [December 26, 2016, 1:43pm UTC](https://discuss.elastic.co/t/es-5-1-server-nodes-ssd-vs-ceph/70008/1 "2016-12-26T13:43:38Z")

</div>

I have read the latest post reagaring this subject, but stil got a question about it, because here is the offers I got for a new server:

> Server 1:  
> 2 vCores  
> 2,4 GHz  
> 8 GB de RAM  
> SSD 40 GB  
> RAID 10 local

and

> Server 2:  
> 1 vCore  
> 2.4 GHz  
> 6 GB RAM  
> 25 GB  
> High Availability (Ceph)

1. We should not ever use Ceph in a node? Why?

2. If we can use Ceph, what would be the better, a server with SSD or with Ceph?

And one last question...

Would Elasticsearch recomend using docker to create new nodes, Kibana etc in the same server?

We can not afford to have another server at the moment, so we will get 1 server to run the ES nodes.

Thanks for any reply on that.

---

<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:** [December 26, 2016, 10:26pm UTC](https://discuss.elastic.co/t/es-5-1-server-nodes-ssd-vs-ceph/70008/2 "2016-12-26T22:26:28Z")

</div>

CEPH is distributed storage, ES is a distributed system.  
Running a distributed system on a distributed FS is asking for latency problems.

---

<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:** [December 26, 2016, 10:26pm UTC](https://discuss.elastic.co/t/es-5-1-server-nodes-ssd-vs-ceph/70008/3 "2016-12-26T22:26:46Z")

</div>

> [@evert](#):
>
> Would Elasticsearch recomend using docker to create new nodes, Kibana etc in the same server?

That's definitely something we see a lot of.

---

<div class="post-metadata">

**Author:** ![evert](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/evert/32/8143_2.png) [@evert](https://discuss.elastic.co/u/evert)\
**Post date:** [December 27, 2016, 10:17am UTC](https://discuss.elastic.co/t/es-5-1-server-nodes-ssd-vs-ceph/70008/4 "2016-12-27T10:17:43Z")

</div>

Thanks @warkolm!

I will get the ssd then with large RAM and put it on live soon.

But the way, I am using docker for the last 8 month in development enviornment and it seems pretty stable. I will get it done in production as well, any further comments I will post here to keep track of it.

Really appreciate your comment.

Happy New Year!

---

<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:** [January 24, 2017, 10:17am UTC](https://discuss.elastic.co/t/es-5-1-server-nodes-ssd-vs-ceph/70008/5 "2017-01-24T10:17:44Z")

</div>

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