Did any one come accross this situation? and any ideas to fix it?
we have Grafana (2 nodes) + ElasticSearch (3 master and 2 data nodes) Integration environment.
ElasticSearch nodes peformance metrics are configured on Grafana.
we have problem recently with ssd disks and it has been fixed. so we tried to add data node 6 back on to cluster, but it still causes heavy cpu iowait time.
any suggestions/tools to investigate and to identify what is the reason for high iowait time?