Do sharding & flush hot threads indicate slow indexing?

Elastic Stack Elasticsearch
1

Hello everyone,

I notice in my hot threads that it's spending a lot of time in write related sections of code. Also, I'm experiencing slower indexing rate than usual (45% decrease)

Here's the output: https://gist.github.com/vinicioflores/fef2cd267224b1532c933eed57b4831c

Please help,

Regards,

2

Hey

Are you running on a slower disk? Is it possible that you are by accident limiting write speeds, so that merging blocks indexing to not write too much data so that merging cannot keep up?

--Alex

3

How would you advice me to check this from ES cluster ? Any metric I can monitor to find out

4

Here's my hot threads output, please help:

I notice that in most of my data nodes, it invests a lot of time in stuck in this method:

org.elasticsearch.common.util.concurrent.ThreadContext$ContextPreservingAbstractRunnable.doRun(ThreadContext.java:672)

5

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