# Maximum Heap Size for Server with 192GB RAM

**URL:** <https://discuss.elastic.co/t/maximum-heap-size-for-server-with-192gb-ram/253277>\
**Category:** Elasticsearch\
**Created:** [October 26, 2020, 8:32am UTC](https://discuss.elastic.co/t/maximum-heap-size-for-server-with-192gb-ram/253277 "2020-10-26T08:32:12Z")\
**Posts on this page:** 7\
**Page:** 1

<div class="post-metadata">

**Author:** ![gnome](https://avatars.discourse-cdn.com/v4/letter/g/87869e/32.png) [@gnome](https://discuss.elastic.co/u/gnome)\
**Post date:** [October 26, 2020, 8:32am UTC](https://discuss.elastic.co/t/maximum-heap-size-for-server-with-192gb-ram/253277/1 "2020-10-26T08:32:12Z")

</div>

Good day everyone. I spend some time read up the maximum heap size can be allocated for ElasticSearch Server, most of it said recommend maximum of 32GB or 1/2 of your physical memory.

I have a server with 192GB of Ram, can I allocate the heap size of 96GB? or maximum support at 30GB?

---

<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:** [October 26, 2020, 2:44pm UTC](https://discuss.elastic.co/t/maximum-heap-size-for-server-with-192gb-ram/253277/2 "2020-10-26T14:44:22Z")

</div>

It's not recommended to go above 30gb.  
With that size of RAM you can think of running 2 data nodes of 30gb each and 1 master node with something like 8gb of heap.

IMHO.

---

<div class="post-metadata">

**Author:** ![gnome](https://avatars.discourse-cdn.com/v4/letter/g/87869e/32.png) [@gnome](https://discuss.elastic.co/u/gnome)\
**Post date:** [October 26, 2020, 3:00pm UTC](https://discuss.elastic.co/t/maximum-heap-size-for-server-with-192gb-ram/253277/3 "2020-10-26T15:00:26Z")

</div>

Thanks for the info.

I think I will convert the machine into KVM to run 2 nodes ES. Currently it's a single node ES machine, no cluster.

---

<div class="post-metadata">

**Author:** ![gnome](https://avatars.discourse-cdn.com/v4/letter/g/87869e/32.png) [@gnome](https://discuss.elastic.co/u/gnome)\
**Post date:** [October 27, 2020, 11:13am UTC](https://discuss.elastic.co/t/maximum-heap-size-for-server-with-192gb-ram/253277/4 "2020-10-27T11:13:02Z")

</div>

My index is create by date, daily. i.e data-20201027, each index is around 300GB of size.

Should I create like 10 shards and 1 replica (to maintain 30GB per shard) on the 2 data nodes?

---

<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:** [October 27, 2020, 2:54pm UTC](https://discuss.elastic.co/t/maximum-heap-size-for-server-with-192gb-ram/253277/5 "2020-10-27T14:54:17Z")

</div>

May be with that size of data, you can have only one node per machine. The RAM will be used for the file system cache.

---

<div class="post-metadata">

**Author:** ![gnome](https://avatars.discourse-cdn.com/v4/letter/g/87869e/32.png) [@gnome](https://discuss.elastic.co/u/gnome)\
**Post date:** [October 28, 2020, 5:57am UTC](https://discuss.elastic.co/t/maximum-heap-size-for-server-with-192gb-ram/253277/6 "2020-10-28T05:57:18Z")

</div>

Thanks. Let me try out multiple node and single node, to find out the different. 🙂 Thanks.

---

<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:** [November 25, 2020, 5:57am UTC](https://discuss.elastic.co/t/maximum-heap-size-for-server-with-192gb-ram/253277/7 "2020-11-25T05:57:20Z")

</div>

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