Hi ,
ES=1.7.5
From time to we keep getting GC Old gen (we are working on ways to minimize it ) , ES trigger long cluster re balance . Anyway to avoid cluster re balance on such event ?
Please take a look at this documentation which covers this question:
https://www.elastic.co/guide/en/elasticsearch/reference/1.7/delayed-allocation.html
And the real solution is as you mentioned to minimise long GCs as much as possible (https://www.elastic.co/guide/en/elasticsearch/guide/current/_monitoring_individual_nodes.html)
Hi ,
I already try :
curl -XPUT 'localhost:9200/_all/_settings' -d '{
"settings": {
"index.unassigned.node_left.delayed_timeout": "5m"
}
}'
And still seeing long re balance , i consider setting cluster.routing.allocation.balance.threshold" : "100.0f , but it will stop the balancer
This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.