A telemetry ingestion pipeline collects data from 100,000 IoT sensors. Each sensor transmi… — AWS Certified Solutions Architect - Associate

A telemetry ingestion pipeline collects data from $100,000$ IoT sensors. Each sensor transmits a $1.5\text{ KB}$ payload every $10\text{ seconds}$ to an Amazon Kinesis data stream configured with $10$ provisioned shards. During peak periods, downstream consumers experience data gaps, and the sensor applications log frequent `ProvisionedThroughputExceededException` errors. Which action should a solutions architect take to resolve this issue with the least operational overhead?

Change the capacity mode of the Kinesis data stream to On-Demand.Answer
B
Increase the number of provisioned shards in the Kinesis data stream to $12$ .
C
Migrate the ingestion layer from Amazon Kinesis Data Streams to a standard Amazon SQS queue to buffer the telemetry spikes.
D
Configure an AWS Lambda function triggered by an Amazon CloudWatch alarm to programmatically scale the stream's shard count when utilization exceeds $80\%$ .

Answer

Change the capacity mode of the Kinesis data stream to On-Demand.

The correct option is to change the capacity mode of the Kinesis data stream to On-Demand. The total ingestion requirements are

10,000\text{ records/s}

and

15\text{ MB/s}

of throughput. Since a single shard is limited to

1\text{ MB/s}

of write throughput, the stream requires a minimum of

15

shards. The current configuration of

10

shards is insufficient. Switching to On-Demand capacity mode enables automatic scaling of shards to accommodate the peak write rate, eliminating the ProvisionedThroughputExceededException errors with minimal administrative effort.

Step-by-Step Solution

Calculate the peak throughput and record rate of the IoT ingestion pipeline.

Record Rate =

10,000\text{ records/s}

. Throughput =

10,000\text{ records/s} \times 1.5\text{ KB} = 15,000\text{ KB/s} = 15\text{ MB/s}

To determine the minimum shard capacity required for ingestion.

Compare the calculated requirement against the current stream capacity.

Current capacity is

10\text{ MB/s}

and

10,000\text{ records/s}

. The required write throughput (

15\text{ MB/s}

) exceeds the provisioned capacity limit (

10\text{ MB/s}

To verify if the write throughput limit is being exceeded, causing the ProvisionedThroughputExceededException.

Choose the solution that resolves the limitation with the lowest operational overhead.

Select On-Demand capacity mode instead of manually resizing the stream or writing custom scaling automation.

On-Demand capacity mode requires zero administrative management and adapts dynamically to traffic peaks.

Key Concept

Kinesis Data Streams Capacity Planning and Auto-Scaling

Estimated Time:1m 30s

Question

Answer

Step-by-Step Solution

Key Concept

Common Mistakes

Topics

More questions from this topic