I'm wondering what will be the differences between a Cassandra cluster with a few large, powerful nodes, and a cluster with many small nodes. For example, if I choose to run Cassandra on AWS EC2, what will be the difference if I choose one m4.4xlarge node, or two m4.2xlarge nodes, or 4 m4.xlarge, or 8 m4.large? They should have the same compute power, according to https://aws.amazon.com/ec2/pricing/. But will they have different performance in reality for Cassandra? Assume we always set replication factor as 1 so the comparison is fair.
1 Answers
Related