Tuning thread pool write queue size

Hello everyone. I was wonder about increase the thread pool for the write queue size.
My error was:

EsThreadPoolExecutor[name = es-data-node/write, queue capacity = 200, org.elasticsearch.common.util.concurrent.EsThreadPoolExecutor@18e72b5a[Running, pool size = 16, active threads = 16, queued tasks = 201, completed tasks = 111090581]]]"},"status":429}

Those errors above can relate to many things: disk slow, low queue…

This data node has 16 core CPUs and 32Gb Ram.
Queued task original was 200, I increased it to a little higher to prevent this error appear
as long time I do at google but I still not find my answer at all about change thread write pool size. Original is 16 thread pool for write ( as I understand it based CPU core, but at the time it used full 16 threads, my CPU usage seem like used 20% only. )
So I think to change it to a little higher will not affect at all.
But i want to discuss about this first. Anyone have done this before? :smiley:

@BlackMetalz Did you found anyway?
I am facing the same issue and I can see we have to change or add some configs in
elasticsearch.yml/opensearch.yml

but didn’t find a way to edit or add configs to opensearch.yml while spinning up a cluster in AWS

some related links
https://aws.amazon.com/premiumsupport/knowledge-center/opensearch-resolve-429-error/
https://www.elastic.co/guide/en/elasticsearch/reference/7.16/modules-threadpool.html#fixed-auto-queue-size

not really, temp solution was only increasing queue, if I use HDD I would increase to 3000 if SSD it can go double or triple depending on the load