I have been trying to reduce our swapping:
The cluster is running on Azure VMs (ES is actually running on a secondary hd):
On one of the nodes:
-
I have ES_HEAP_SIZE set to 14g
-
I have the Windows Virtual Memory page file disabled
-
I have mlock enabled
"aVL-ZYsnQju5G4ZsVVwT5Q": { "settings": { "bootstrap": { "mlockall": "true" } }, "process": { "mlockall": true } },
I still have these warnings about swapping. Note that I have not rebooted the VM. I don't want the node to be down that long. Is that required?
Any idea how to tune this?