# Heavy Bulk inserts -Read & Write performance

**URL:** <https://discuss.elastic.co/t/heavy-bulk-inserts-read-write-performance/284107>\
**Category:** Elasticsearch\
**Created:** [September 13, 2021, 4:42pm UTC](https://discuss.elastic.co/t/heavy-bulk-inserts-read-write-performance/284107 "2021-09-13T16:42:19Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Vishnu\_Singhal](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/vishnu_singhal/32/52488_2.png) [@Vishnu\_Singhal](https://discuss.elastic.co/u/Vishnu_Singhal)\
**Post date:** [September 13, 2021, 4:42pm UTC](https://discuss.elastic.co/t/heavy-bulk-inserts-read-write-performance/284107/1 "2021-09-13T16:42:19Z")

</div>

Hi,  
Im using ES 7.14.0 with 16 data hot nodes and 2 warm nodes. When I'm doing heavy inserts, there is high thread pool write active count, and sometimes I get NONodeLeftException  
I'm using ES-Hadoop Library with spark 3.1.0 on kuberenetes.

If i have made index refresh interval to -1 jobs are not having any issues. BUt I make it -1 documents will not be searchable.

Also tried refresh\_interval with 600s to 100s but not much help, the probability of error reduces but jobs takes time.

How can i achieve heavy write & read from elasticsearch ?

 ![Screenshot from 2021-09-13 20-35-48](https://us1.discourse-cdn.com/elastic/original/3X/1/d/1d925e8d9955c1d9faae4aed3bb1d697f905fa44.png)

---

<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 11, 2021, 4:43pm UTC](https://discuss.elastic.co/t/heavy-bulk-inserts-read-write-performance/284107/2 "2021-10-11T16:43:18Z")

</div>

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