This question considers a provisioned capacity with autoscaling. This is a very basic question to better understand WCU/RCU pricing.
I'm not quite sure of the answer because it appears that provisioning a high throughput for a short period might be more expensive (i.e. require more resource reservation) than provisioning a low throghput in a long period of time.
I have a huge workload to import into a DynamoDB table (~1 billion records) and I want to understand if I could increase the throughput to reduce the time needed for the operation to finish, understanding if that would incur in the same price.
Example: will using 5000 WCU for 1 minute incur the same cost as using 5 WCU for 1000 minutes?
This question seems to be very similar but I still don't understand.
The cost is per WCU-hour with billing per hour. Provisioning 100,000 for 5 hours or 50,000 for 10 hours results in the same WCU-hr consumption.