# Tipps for a Cluster setup

**URL:** <https://discuss.elastic.co/t/tipps-for-a-cluster-setup/38715>\
**Category:** Elasticsearch\
**Created:** [January 8, 2016, 10:57am UTC](https://discuss.elastic.co/t/tipps-for-a-cluster-setup/38715 "2016-01-08T10:57:28Z")\
**Posts on this page:** 5\
**Page:** 1

<div class="post-metadata">

**Author:** ![Michael\_\_Rennecke](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/michael__rennecke/32/3484_2.png) [@Michael\_\_Rennecke](https://discuss.elastic.co/u/Michael__Rennecke)\
**Post date:** [January 8, 2016, 10:57am UTC](https://discuss.elastic.co/t/tipps-for-a-cluster-setup/38715/1 "2016-01-08T10:57:28Z")

</div>

Hello,

I will setup a Elasticsearchcluster for logging (Elasticsearch, Logstash Kibana). I have 2 "old" Database servers with 128 GB RAM 2x 12 core 4x 3.4 TB SAS 7200 rpm and 700 GB PCI-X-SSD. What is the best setup for this hardware?

1. (2 nodes per server) RAID 5 with all disks (ext4 or xfs), SSD as bcache device
2. (2 nodes per server) zfs and the SSD as L2ARC and ZIL
3. hot-warm-architecture: one node on the SSD for the daily logs and one node on a RAID 5

The RAID 5 is for the redundancy of the OS.

cheers,  
Michael

---

<div class="post-metadata">

**Author:** ![dadoonet](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/dadoonet/32/137187_2.png) [@dadoonet](https://discuss.elastic.co/u/dadoonet)\
**Post date:** [January 8, 2016, 11:11am UTC](https://discuss.elastic.co/t/tipps-for-a-cluster-setup/38715/2 "2016-01-08T11:11:53Z")

</div>

I like the hot-warm architecture.

BTW I'd also add a master node only on the warm machine with 2 gb RAM, make the hot nodes not master elligible.

---

<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:** [January 9, 2016, 10:18pm UTC](https://discuss.elastic.co/t/tipps-for-a-cluster-setup/38715/3 "2016-01-09T22:18:00Z")

</div>

> [@Michael\_\_Rennecke](#):
>
> The RAID 5 is for the redundancy of the OS.

It's also slow, which may impact your ES performance.

---

<div class="post-metadata">

**Author:** ![Michael\_\_Rennecke](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/michael__rennecke/32/3484_2.png) [@Michael\_\_Rennecke](https://discuss.elastic.co/u/Michael__Rennecke)\
**Post date:** [January 11, 2016, 7:06am UTC](https://discuss.elastic.co/t/tipps-for-a-cluster-setup/38715/4 "2016-01-11T07:06:26Z")

</div>

I know, but 4 disks is a bad number. The other option is one mirror and 2 disks.This Setup sound broken. A redundant OS is a request from my boss.

---

<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 5, 2017, 11:25pm UTC](https://discuss.elastic.co/t/tipps-for-a-cluster-setup/38715/5 "2017-07-05T23:25:32Z")

</div>