Hello we have a elasticsearch(5.6) cluster and due to development overhead we can not upgrade our code for newer version of elasticsearch near soon.
So there are some issues we are facing with it. Numbers one is high CPU consumption, After running the given command for hod thread below, it's showing big Integer conversion. where average CPU consumption is 90%+.
curl http:/localhost:9200/_nodes/hot_threads?human
Strange behavior:
When ever out of 3 nodes we restart one node(node-2) consuming high cpu, other nodes start getting high CPU when we restart high gc node(node-2).
Also it's the same node(node-2) which gets high cpu every week. so questions is are the other 2 nodes not getting any similar search thread requests. that are causing high cpu on node-2
Thanks