Maybe i have missed something. But haven't found anything well documented about insert mechanics.
As i understand, insert is synchronous operation and it's inserting new data to all data nodes related to this partition.
So basically how much replicas out there of partition, then that much slower insert operations will become?
Replicas amount directly affecting performance/speed of insert/update/delete operations?
So only read/select operations are scaling? But with each new servers added in the cluster, write operations will become slower.
Or i don't understand something...
Thank you.
