Amazon Kinesis Data Analytics pricing

With Amazon Kinesis Data Analytics, you pay only for what you use. You are charged an hourly rate based on the average number of Kinesis Processing Units (or KPUs) used to run your stream processing application.

A single KPU is a unit of stream processing capacity comprised of 4 GB memory, 1 vCPU compute, and corresponding networking capabilities. As the complexity of your queries varies, and the demands on memory and compute vary in response, Amazon Kinesis Data Analytics will automatically and elastically scale the number of KPUs required to complete your analysis. There are no resources to provision, no upfront costs, or minimum fees associated with Amazon Kinesis Data Analytics.

Simple Streaming Filter

A customer uses Amazon Kinesis Data Analytics to continuously filter streaming data captured by their Amazon Kinesis Data Firehose and only retain records of interest. Their monthly Amazon Kinesis Data Analytics charges will be computed as follows:

Monthly charges

The price in US-East is $0.11 per KPU-Hour used for their stream processing application. 1 KPU provides stream processing capacity equivalent to 4 GB memory, 1 vCPU compute, and corresponding networking capabilities.

This simple streaming filter consumes 1.7 GB of memory and 50% of 1 vCPU to process the data in the Firehose.

Since their streaming SQL query uses less than 4GB of memory and less than 1 vCPU of compute, their application will only require 1 KPU to process the workload.

Monthly Charges = 30 Days * 24 Hours * (1 KPU * $0.11/ Hour) = $79.20

Sliding Window With Workload Spikes

A customer uses Amazon Kinesis Data Analytics to compute a 1-minute, sliding-window sum of items sold in online shopping transactions captured in their Kinesis Stream. This Stream normally ingests data at 1,000 records/ second but the data spikes once a day during promotional campaigns to 6,000 records/ second inside an hour. Their monthly Amazon Kinesis Data Analytics charges will be computed as follows:

Monthly charges

The price in US-East is $0.11 per KPU-Hour. 1 KPU provides stream processing capacity equivalent to 4 GB memory, 1 vCPU compute, and corresponding networking capabilities.

The incoming Kinesis data stream transmits data at 1,000 records/ second. However once a day, inside an hour, the Stream spikes to 6,000 records/ second.

• For ‘steady state’ which occurs 23 of 24 hours in the day:

The sliding-window query consumes 3.2 GB of memory and 50% of 1 vCPU to process the data.

Since their streaming SQL query uses less than 4GB of memory and less than 1 vCPU of compute, their application will only require 1 KPU to process the workload during these hours.

• For ‘spiked state’ which occurs for 1 of 24 hours in the day:

The sliding-window query consumes 6.4 GB of memory and 75% of 1 vCPU to process the higher data throughput in the stream. The streaming SQL query exceeds the memory provided by a single KPU, but is less than 8 GB which is the amount of memory provided by 2 KPUs. The customer will be billed for 2 KPUs for that 1 hour out of the 24 hours in the day.

Monthly Charges:

Steady State = 30 Days * 23 Hours * (1 KPU * $0.11/ Hour) = $75.90

Spiked State = 30 Days * 1 Hour * (2 KPUs * $0.11/ Hour) = $6.60

Total Charges = $75.90 + $6.60 = $82.50