I have a 3 node elastic cluster backing my graylog. I have replication factor 1 turned on, on all indexes. When rebooting any single elastic node my cluster should go to a yellow state. But recently it seems that rebooting an elastic node is causing my cluster to go red (for a little while).
Is there something I mis configured or missed that would cause this to happen?
Does the number of indexes and or index sets have any impact on replication and node count?
Is three nodes with 1084 active shards insufficient , do I need a higher replication factor?
I am running graylog 3.3.4-1 on ubuntu 16.04
with elastic 6.8.11 and mongodb 4.0.19