We have currently a project running where we need to store 1 to 20,30 million 500B to 2kB records.
We need to use MySQL cluster on 4 EC2 instances (8GB Ram, 2 cores) and ephemeral storage (Records don't need to be stored for a long time). Our client application runs also on EC2.
What latency for read and update ops can we expect circa?
UPDATE: WE have executed some tests with one SQL Node and three Data Nodes. The results are just poor. Data retrieval tooks seconds. Can you imagine why MySQL cluster does perform so poor on EC2 m1.large?