The funny thing all about AWS/ES drama is AWS Elasticsearch is 1.5x price. It puts a stupid ELB-load balancer that cannot detect timely if the backend services

The funny thing all about AWS/ES drama is AWS Elasticsearch is 1.5x price. It puts a stupid ELB-load balancer that cannot detect timely if the backend services are unhealthy. Updates don't provide rolling update capability, its always blue green. Updates takes hours depending on your data size. And during this shuffle, your CPU usages are top. The nodes get down for inexplicable reasons and not replaced for many minutes. Your retries don't cut it because of that non-smart load balancer. It limits the disk size by instance types so you have no option but to scale horizontally.I just want bigger disks, you don't know my indicies, how I use them, don't put arbitrary limits. It's just a convenience for procurement and VPCs, nothing more.