Discover: Request Timeout after 30000ms: ES

[rahulnama]

Hello Team

Greetings of the day!

I'm using ES on a node.

Node Specifications:
2 VCPUs, 50GB DISK, 8 GB RAM
Indices 4
Every Indice has less than 100mb of data.
When I'm trying to visualise an indice data in discover, it's displaying Request Timeout error as CPU is being utilized 100%. I have 10 shards for that particular indice.
Is there anything that I could to here to avoid this? I read I can increase the timeout but doing so will help me? is it a best practice?

Please do suggest. Thanks for your time as always

Rahul Nama

[thiago (Thiago Souza)]

How much heap is set to JVM? Also, how many shards are there is this node (considering all open indices)?

[rahulnama]

Hello @thiago

I have 12 shards in this node. and heap memory set to jvm is 4GB.

[thiago (Thiago Souza)]

So 2 VCPUs is not enough for that. We recommend at least for 4 cores. You can check Nodes hot_threads to get a better understanding of what's burning CPU.

[Christian_Dahlqvist]

What is the average size of your documents? You can try disabling highlighting by setting doc_table:highlight to false under the advanced options tab as this can reduce CPU usage quite a bit.

[rahulnama]

@Christian_Dahlqvist

Size of Documents is very less(in kbs). Sure I will do that and give one more try

Thank you