High CPU Utilization - Tune performance

Elastic Stack Elasticsearch
1

Hi team,

First time posting a topic in here.

I have a 6-node cluster which look like this:

heap.percent ram.percent cpu load_1m load_5m load_15m node.role master name
          31          92  18    9.53    8.86     8.77 dimr      -      node-6
          26          93  29   15.48   18.94    25.95 dimr      *      node-5
          71          91  25   14.23   17.64    25.82 dimr      -      node-3
          34          71  80   40.09   32.82    23.56 dimr      -      node-2
          68          83  21   11.65   10.55    10.31 dimr      -      node-4
          30          79  92   36.20   30.51    21.76 dimr      -      node-1

As you can see, 2 nodes have high CPU usage and per my observation it rotates. There's always 2 nodes within the cluster that has high CPU utilization. Usually above 70% and gets as high as 90%+

Based on initial checking, data stream logs-system.security [Fleet-Managed] has the most indices with more than 22,000/s index rate(based on Stack Monitoring - Indices tab). Default configuration for this is to have 1 primary shard with default rollover of 50gb maximum primary shard size.

Appreciate any help I can get on this issue.

Thanks

2

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