Hi Elastic team,
I have deployed Elasticsearch 6.x in GKE and I want to perform node pool upgrade. I have checked this doc https://www.elastic.co/guide/en/elasticsearch/reference/current/rolling-upgrades.html
I will create ES nodes on the new machine, stop all balancing and reallocation, and will reroute all data shards to new ES nodes.
My question is that can I migrate all data shards at once without any downtime ( or increase in search fetch time ) or should I do it one by one or any other number.
Thanks for your support.