I am using ES 7.10.1. My cluster has 20 nodes.
I would start these 20 nodes one by one (the cluster already has data). During the startup, would rebalance happen because some have started up while some haven't?
I would if rebalance may happen, how could I avoid this? Thanks.
In order to prevent rebalancing, you may run the following command, which is typically what is run before doing a rolling upgrade in order to prevent shards from being rebalanced to other nodes.
PUT _cluster/settings
{
"persistent": {
"cluster.routing.allocation.enable": "none"
}
}