Elasticsearch cluster and 2 servers went down

Hi Team,

I have 6 server cluster with 5 data node and 1 coordinating node , 2 of the server went completely down because of the heavy load.

On one server even telnet is not working on 22 port and on the other sever when I am trying to connect, I am getting Network error: Software Caused connection abort.

My cluster is red and logstash is not picking message on most of the index.

I have 5 shards and 1 replica for each index. Is there is a way to resolve it.

Can anyone please help.

One of the error, I can see is as below

Very big device: Trying to use Read capacity(16)

What version are you on?
What do the Elasticsearch logs show?