# Disk I/O for master-only node?

**URL:** <https://discuss.elastic.co/t/disk-i-o-for-master-only-node/149365>\
**Category:** Elasticsearch\
**Created:** [September 20, 2018, 10:18pm UTC](https://discuss.elastic.co/t/disk-i-o-for-master-only-node/149365 "2018-09-20T22:18:51Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![YuWatanabe](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/yuwatanabe/32/13259_2.png) [@YuWatanabe](https://discuss.elastic.co/u/YuWatanabe)\
**Post date:** [September 20, 2018, 10:18pm UTC](https://discuss.elastic.co/t/disk-i-o-for-master-only-node/149365/1 "2018-09-20T22:18:51Z")

</div>

Hello.

I would like to get advice for hardware requirements for dedicated-master node with es v 6.3.2.  
This is my first time setting up dediciated master-node so I appreciate if I could get advice.

I have read below 2 articles but I was not sure about how frequent the cluster state is update to the disk.

[Master-eligible Node](https://www.elastic.co/guide/en/elasticsearch/reference/current/modules-node.html#master-node)

[What is the ES cluster size that requiers a dedicated master node?](https://discuss.elastic.co/t/what-is-the-es-cluster-size-that-requiers-a-dedicated-master-node/26489)

Currently, I am building new cluster and planning to set up 3 dedicated-master node with below hw. Indexing rate for data nodes are estimated up to _50GB/day_ using _5 data nodes_.

1. CPU: Xeon Bronze 3106 (8 cores , 1.7 GHz)
2. RAM: 32 GB
3. HDD: 300 GB SAS 15000RPM \*2 (RAID1)

For RAM , I see that 4GB of heap is a starting line but planning to have little bit of room to scale up the heap size just in case. However for _disks_ if cluster state is updated in some interval, is spinning-disk enough ? Or if cluster state is updated very frequently , should you use ssd?

Best Regards,  
Yu Watanabe

---

<div class="post-metadata">

**Author:** ![nik9000](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/nik9000/32/44947_2.png) [@nik9000](https://discuss.elastic.co/u/nik9000)\
**Post date:** [September 20, 2018, 10:31pm UTC](https://discuss.elastic.co/t/disk-i-o-for-master-only-node/149365/2 "2018-09-20T22:31:32Z")

</div>

The cluster state is updated when indices are created and when shards move from node to node and at times like that. But the cluster state is comparatively small so a spinning disk should be just fine.

---

<div class="post-metadata">

**Author:** ![YuWatanabe](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/yuwatanabe/32/13259_2.png) [@YuWatanabe](https://discuss.elastic.co/u/YuWatanabe)\
**Post date:** [September 21, 2018, 7:41am UTC](https://discuss.elastic.co/t/disk-i-o-for-master-only-node/149365/3 "2018-09-21T07:41:41Z")

</div>

@nik9000

Thank you for the useful advice.

---

<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 19, 2018, 7:50am UTC](https://discuss.elastic.co/t/disk-i-o-for-master-only-node/149365/4 "2018-10-19T07:50:17Z")

</div>

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