Kinesis Data Streams vs Kinesis Data Firehose

AWS Kinesis Data Streams vs Kinesis Data Firehose

Kinesis Data Streams vs. Kinesis Data Firehose
~ Last updated on : ~ jayendrapatil ~ 11 Comments

Kinesis Data Streams vs. Firehose

Purpose

  • Kinesis data streams is highly customizable and best suited for developers building custom applications or streaming data for specialized needs.
  • Kinesis Data Firehose handles loading data streams directly into AWS products for processing. Firehose also allows for streaming to S3, OpenSearch Service, Redshift, Apache Iceberg tables, Snowflake, and other destinations, where data can be copied for processing through additional services.

Provisioning & Scaling

  • Kinesis Data Streams offers two capacity modes:
    • Provisioned Mode: Requires manual configuration of shards and scaling. You specify the number of shards needed based on expected throughput.
    • On-Demand Mode (Launched November 2021): Automatically scales to handle gigabytes of write and read throughput per minute without manual shard management. Default capacity of 4 MB/s write (4000 records/s), can scale up to 200 MB/s (or 1 GB/s with limit increase).
  • Kinesis Data Firehose is fully managed and sends data to S3, Redshift, OpenSearch, Apache Iceberg tables, Snowflake, and other destinations. Scaling is handled automatically, up to gigabytes per second, and allows for batching, encrypting, and compressing.

Processing Delay

  • Kinesis Data Streams provides real-time processing with ~200 ms for shared throughput classic single consumer and ~70 ms for the enhanced fan-out consumer.
  • Kinesis Data Firehose provides near real-time processing:
    • Standard Buffering: Minimum buffer time of 60 seconds (1 min), maximum 900 seconds (15 min)
    • Zero Buffering (Announced December 2023): Delivers data within ~5 seconds with no buffering delay, enabling real-time use cases

Data Storage

  • Kinesis Data Streams provide data storage. Data typically is made available in a stream for 24 hours, but for an additional cost, users can gain data availability for up to 365 days (8760 hours).
  • Kinesis Data Firehose does not provide data storage.

Replay

  • Kinesis Data Streams supports replay capability
  • Kinesis Data Firehose does not support replay capability

Producers & Consumers

  • Kinesis Data Streams & Kinesis Data Firehose support multiple producer options including SDK, KPL, Kinesis Agent, IoT, etc.
  • Kinesis Data Streams support multiple consumers option including SDK, KCL, and Lambda, and can write data to multiple destinations. However, they have to be coded.
  • Kinesis Data Firehose consumers are close-ended and support destinations including:
    • Amazon S3
    • Amazon Redshift
    • Amazon OpenSearch Service
    • Amazon OpenSearch Serverless
    • Apache Iceberg Tables (Added October 2024) – Stream data directly into Iceberg format tables in S3
    • Snowflake (with Snowpipe Streaming) – Real-time streaming to Snowflake
    • Splunk
    • Third-party HTTP endpoints (Datadog, Dynatrace, New Relic, MongoDB, Coralogix, Elastic, etc.)

Key Differences Summary

Feature Kinesis Data Streams Kinesis Data Firehose
Capacity Mode Provisioned or On-Demand Fully managed (automatic)
Latency Real-time (~70-200 ms) Near real-time (60s default, ~5s with zero buffering)
Data Retention 24 hours to 365 days No storage
Replay ✅ Supported ❌ Not supported
Consumers Custom (SDK, KCL, Lambda) Pre-defined (S3, Redshift, OpenSearch, Iceberg, Snowflake, etc.)
Use Case Custom processing, real-time analytics ETL, loading to data stores

AWS Certification Exam Practice Questions

  • Questions are collected from Internet and the answers are marked as per my knowledge and understanding (which might differ with yours).
  • AWS services are updated everyday and both the answers and questions might be outdated soon, so research accordingly.
  • AWS exam questions are not updated to keep up the pace with AWS updates, so even if the underlying feature has changed the question might not be updated
  • Open to further feedback, discussion and correction.
  • Questions are collected from Internet and the answers are marked as per my knowledge and understanding (which might differ with yours).
  • AWS services are updated everyday and both the answers and questions might be outdated soon, so research accordingly.
  • AWS exam questions are not updated to keep up the pace with AWS updates, so even if the underlying feature has changed the question might not be updated
  • Open to further feedback, discussion and correction.
  1. Your organization needs to ingest a big data stream into its data lake on Amazon S3. The data may stream in at a rate of hundreds of megabytes per second. What AWS service will accomplish the goal with the least amount of management?
    1. Amazon Kinesis Firehose
    2. Amazon Kinesis Streams
    3. Amazon CloudFront
    4. Amazon SQS
  2. Your organization is looking for a solution that can help the business with streaming data several services will require access to read and process the same stream concurrently. What AWS service meets the business requirements?
    1. Amazon Kinesis Firehose
    2. Amazon Kinesis Streams
    3. Amazon CloudFront
    4. Amazon SQS
  3. Your application generates a 1 KB JSON payload that needs to be queued and delivered to EC2 instances for applications. At the end of the day, the application needs to replay the data for the past 24 hours. In the near future, you also need the ability for other multiple EC2 applications to consume the same stream concurrently. What is the best solution for this?
    1. Kinesis Data Streams
    2. Kinesis Firehose
    3. SNS
    4. SQS
  4. A company needs to stream data to Amazon S3 with the lowest possible latency (under 10 seconds). Which Kinesis service and configuration should they use? (Assume December 2023 or later)
    1. Kinesis Data Streams with Lambda consumer
    2. Kinesis Data Firehose with zero buffering enabled
    3. Kinesis Data Firehose with 60-second buffer
    4. Kinesis Data Streams with KCL consumer
  5. A company wants to avoid manual shard management for their Kinesis Data Streams. Which capacity mode should they use? (Assume November 2021 or later)
    1. Provisioned mode with Auto Scaling
    2. On-Demand mode
    3. Enhanced fan-out mode
    4. Standard mode
  6. A data analytics team needs to stream real-time data into Apache Iceberg tables in S3 for analytics. Which AWS service supports this natively? (Assume October 2024 or later)
    1. Kinesis Data Streams
    2. Kinesis Data Firehose
    3. AWS Glue Streaming
    4. Amazon MSK

References

AWS Kinesis Data Streams vs Kinesis Data Firehose

  • Kinesis acts as a highly available conduit to stream messages between data producers and data consumers.
  • Data producers can be almost any source of data: system or web log data, social network data, financial trading information, geospatial data, mobile app data, or telemetry from connected IoT devices.
  • Data consumers will typically fall into the category of data processing and storage applications such as Apache Hadoop, Apache Storm, S3, and ElasticSearch.

Kinesis Data Streams vs. Firehose

Purpose

  • Kinesis data streams is highly customizable and best suited for developers building custom applications or streaming data for specialized needs.
  • Kinesis Data Firehose handles loading data streams directly into AWS products for processing. Firehose also allows for streaming to S3, OpenSearch Service, Redshift, Apache Iceberg tables, Snowflake, and other destinations, where data can be copied for processing through additional services.

Provisioning & Scaling

  • Kinesis Data Streams offers two capacity modes:
    • Provisioned Mode: Requires manual configuration of shards and scaling. You specify the number of shards needed based on expected throughput.
    • On-Demand Mode (Launched November 2021): Automatically scales to handle gigabytes of write and read throughput per minute without manual shard management. Default capacity of 4 MB/s write (4000 records/s), can scale up to 200 MB/s (or 1 GB/s with limit increase).
  • Kinesis Data Firehose is fully managed and sends data to S3, Redshift, OpenSearch, Apache Iceberg tables, Snowflake, and other destinations. Scaling is handled automatically, up to gigabytes per second, and allows for batching, encrypting, and compressing.

Processing Delay

  • Kinesis Data Streams provides real-time processing with ~200 ms for shared throughput classic single consumer and ~70 ms for the enhanced fan-out consumer.
  • Kinesis Data Firehose provides near real-time processing:
    • Standard Buffering: Minimum buffer time of 60 seconds (1 min), maximum 900 seconds (15 min)
    • Zero Buffering (Announced December 2023): Delivers data within ~5 seconds with no buffering delay, enabling real-time use cases

Data Storage

  • Kinesis Data Streams provide data storage. Data typically is made available in a stream for 24 hours, but for an additional cost, users can gain data availability for up to 365 days (8760 hours).
  • Kinesis Data Firehose does not provide data storage.

Replay

  • Kinesis Data Streams supports replay capability
  • Kinesis Data Firehose does not support replay capability

Producers & Consumers

  • Kinesis Data Streams & Kinesis Data Firehose support multiple producer options including SDK, KPL, Kinesis Agent, IoT, etc.
  • Kinesis Data Streams support multiple consumers option including SDK, KCL, and Lambda, and can write data to multiple destinations. However, they have to be coded.
  • Kinesis Data Firehose consumers are close-ended and support destinations including:
    • Amazon S3
    • Amazon Redshift
    • Amazon OpenSearch Service
    • Amazon OpenSearch Serverless
    • Apache Iceberg Tables (Added October 2024) – Stream data directly into Iceberg format tables in S3
    • Snowflake (with Snowpipe Streaming) – Real-time streaming to Snowflake
    • Splunk
    • Third-party HTTP endpoints (Datadog, Dynatrace, New Relic, MongoDB, Coralogix, Elastic, etc.)

Key Differences Summary

Feature Kinesis Data Streams Kinesis Data Firehose
Capacity Mode Provisioned or On-Demand Fully managed (automatic)
Latency Real-time (~70-200 ms) Near real-time (60s default, ~5s with zero buffering)
Data Retention 24 hours to 365 days No storage
Replay ✅ Supported ❌ Not supported
Consumers Custom (SDK, KCL, Lambda) Pre-defined (S3, Redshift, OpenSearch, Iceberg, Snowflake, etc.)
Use Case Custom processing, real-time analytics ETL, loading to data stores

AWS Certification Exam Practice Questions

  • Questions are collected from Internet and the answers are marked as per my knowledge and understanding (which might differ with yours).
  • AWS services are updated everyday and both the answers and questions might be outdated soon, so research accordingly.
  • AWS exam questions are not updated to keep up the pace with AWS updates, so even if the underlying feature has changed the question might not be updated
  • Open to further feedback, discussion and correction.
  1. Your organization needs to ingest a big data stream into its data lake on Amazon S3. The data may stream in at a rate of hundreds of megabytes per second. What AWS service will accomplish the goal with the least amount of management?
    1. Amazon Kinesis Firehose
    2. Amazon Kinesis Streams
    3. Amazon CloudFront
    4. Amazon SQS
  2. Your organization is looking for a solution that can help the business with streaming data several services will require access to read and process the same stream concurrently. What AWS service meets the business requirements?
    1. Amazon Kinesis Firehose
    2. Amazon Kinesis Streams
    3. Amazon CloudFront
    4. Amazon SQS
  3. Your application generates a 1 KB JSON payload that needs to be queued and delivered to EC2 instances for applications. At the end of the day, the application needs to replay the data for the past 24 hours. In the near future, you also need the ability for other multiple EC2 applications to consume the same stream concurrently. What is the best solution for this?
    1. Kinesis Data Streams
    2. Kinesis Firehose
    3. SNS
    4. SQS
  4. A company needs to stream data to Amazon S3 with the lowest possible latency (under 10 seconds). Which Kinesis service and configuration should they use? (Assume December 2023 or later)
    1. Kinesis Data Streams with Lambda consumer
    2. Kinesis Data Firehose with zero buffering enabled
    3. Kinesis Data Firehose with 60-second buffer
    4. Kinesis Data Streams with KCL consumer
  5. A company wants to avoid manual shard management for their Kinesis Data Streams. Which capacity mode should they use? (Assume November 2021 or later)
    1. Provisioned mode with Auto Scaling
    2. On-Demand mode
    3. Enhanced fan-out mode
    4. Standard mode
  6. A data analytics team needs to stream real-time data into Apache Iceberg tables in S3 for analytics. Which AWS service supports this natively? (Assume October 2024 or later)
    1. Kinesis Data Streams
    2. Kinesis Data Firehose
    3. AWS Glue Streaming
    4. Amazon MSK

References