# Master node on Hot node in hot -warm architecture?

**URL:** <https://discuss.elastic.co/t/master-node-on-hot-node-in-hot-warm-architecture/65909>\
**Category:** Elasticsearch\
**Created:** [November 13, 2016, 1:56pm UTC](https://discuss.elastic.co/t/master-node-on-hot-node-in-hot-warm-architecture/65909 "2016-11-13T13:56:47Z")\
**Posts on this page:** 8\
**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:** [November 13, 2016, 1:56pm UTC](https://discuss.elastic.co/t/master-node-on-hot-node-in-hot-warm-architecture/65909/1 "2016-11-13T13:56:47Z")

</div>

I understand that as it is described in [blog](https://www.elastic.co/blog/hot-warm-architecture) , master node should be deployed as dedicated master nodes. I want to use the warm node to have flexible storage volume with low cost.

However, since I am preparing only 5 nodes, I would like to know if it is possible to build master on hot nodes.

My requirement is ,

Store logs for 1 year  
100 GB events / day  
2500 events / sec

My architecture ,  
node1 hot (SAS (RAID0+1) or SSD(RAID0+1)) + master?  
node2 hot (SAS (RAID0+1) or SSD(RAID0+1)) + master?  
node3 hot (SAS (RAID0+1) or SSD(RAID0+1)) + master?  
node4 warm ( possibility of attaching to external storage)  
node5 warm ( possibility of attaching to external storage)

- I chose external storage since it would be more flexible to increase the disk size

All incoming logs to elasticsearch will come from logstash.

---

<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 13, 2016, 10:50pm UTC](https://discuss.elastic.co/t/master-node-on-hot-node-in-hot-warm-architecture/65909/2 "2016-11-13T22:50:08Z")

</div>

Do you want to run two instances on the same host?

---

<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:** [November 14, 2016, 3:04am UTC](https://discuss.elastic.co/t/master-node-on-hot-node-in-hot-warm-architecture/65909/3 "2016-11-14T03:04:51Z")

</div>

@warkolm

Thanks for the reply. Short answer to your question is No.

So, are hot and warm nodes **dedicated data nodes** and need to set **node.master** to false?

---

<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 14, 2016, 4:03am UTC](https://discuss.elastic.co/t/master-node-on-hot-node-in-hot-warm-architecture/65909/4 "2016-11-14T04:03:29Z")

</div>

If you make the hot nodes master, and you get a big batch of messages and the current hot node that is also master hits high GC, what do you think will happen to the cluster?

---

<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:** [November 14, 2016, 4:10am UTC](https://discuss.elastic.co/t/master-node-on-hot-node-in-hot-warm-architecture/65909/5 "2016-11-14T04:10:05Z")

</div>

Master nodes will have delay in its processing ( creating index, deleting index, track nodes) . Which will have instability with the cluster?

---

<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 14, 2016, 4:35am UTC](https://discuss.elastic.co/t/master-node-on-hot-node-in-hot-warm-architecture/65909/6 "2016-11-14T04:35:25Z")

</div>

Yeah 🙂

Ideally you should have _dedicated_ master nodes, but if not you just need to be aware of the risks you need to balance.

---

<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:** [November 14, 2016, 4:37am UTC](https://discuss.elastic.co/t/master-node-on-hot-node-in-hot-warm-architecture/65909/7 "2016-11-14T04:37:04Z")

</div>

Thanks for the lesson! I like it.

---

<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 12, 2016, 4:37am UTC](https://discuss.elastic.co/t/master-node-on-hot-node-in-hot-warm-architecture/65909/8 "2016-12-12T04:37:32Z")

</div>

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