Data Lakehouse

What Are The Differences Between AWS Redshift Spectrum vs AWS Athena?

Before we begin: Redshift Spectrum vs Redshift

While the thrust of this article is an AWS Redshift Spectrum vs Athena comparison, there can be some confusion with the difference between AWS Redshift Spectrum and AWS Redshift. Very briefly, Redshift is the storage layer/data warehouse. Redshift Spectrum, on the other hand, is an extension to Redshift that is a query engine.

Redshift spectrum

What is Amazon Athena?

Athena is Amazon’s standalone, serverless SQL query engine implementation of Presto. This is used to query data stored on Amazon S3. It is fully managed by Amazon, there is nothing to setup, manage or configure. This also means that the performance can be very inconsistent as you have no dedicated compute resources.

What is Amazon Redshift Spectrum?

Redshift Spectrum is an extension of Amazon Redshift. It is a serverless query engine that can query both AWS S3 data and tabular data in Redshift using SQL. This enables you to join data stored in external object stores with data stored in Redshift to perform more advanced queries.

Key Features & Differences: Redshift vs Athena

Athena and Redshift Spectrum offer similar functionality, namely, serverless query of S3 data using SQL. That makes them easy to manage. This also is more cost-effective as there is nothing to set up and you are only charged based on the amount of data scanned. S3 storage is significantly less expensive than a database on AWS for the same amount of data.

  • Pooled vs allocated resources: Both are serverless, however Spectrum resources are allocated based on your Redshift cluster size. Athena, however, relies on non-dedicated, pooled resources.
  • Cluster management: Spectrum actually does need a bit of cluster management, but Athena is truly serverless.
  • Performance: Performance for Athena depends on your S3 optimization, while Spectrum, as previously noted, depends on your Redshift cluster resources and S3 optimization. If you need a specific query to run more quickly, then you can allocate additional compute resources to it.
  • Standalone vs feature: Redshift Spectrum runs in tandem with Amazon Redshift, while Athena is a standalone query engine for querying data stored in Amazon S3.
  • Consistency: Spectrum provides more consistency in query performance while Athena has inconsistent results due to the pooled resources.
  • Query types: Athena is great for simpler interactive queries, while Spectrum is more oriented towards large, complex queries.
  • Pricing: The cost for both is the same. They run $5 per compressed terabyte scanned, however with Spectrum, you must also consider the Redshift compute costs.
  • Schema management: Both use AWS Glue for schema management, and while Athena is designed to work directly with Glue, Spectrum needs external tables to be configured for each Glue catalog schema.
  • Federated query capabilities: Both support federated queries.

Athena vs Redshift: Functionality

The functionality of each is very similar, namely using standard SQL to query the S3 object store. If you are working with Redshift, then Spectrum can join information in S3 with tables stored in Redshift directly. Athena also has a Redshift connector to allow for similar joins. However if you are using Redshift, it would likely make more sense to use Spectrum in this case.

Athena vs Redshift: Integrations

Keep in mind that when working with S3 objects, these are not traditional databases, which means there are no indexes to be scanned or used for joins. If you are working with files with high-cardinality and trying to join them, you will likely have very poor performance.

When connecting to data sources other than S3, Athena has a connector ecosystem to work with. This system provides a collection of sources that you can directly query with no copy required. Federated queries were added to Spectrum in 2020 and provide a similar capability with the added benefit of being able to perform transformations on the data and load it directly into Redshift tables.

AWS Athena vs Redshift: To Summarize

If you are already using Redshift, then Spectrum makes a lot of sense, but if you are just getting started with the cloud, then the Redshift ecosystem is likely overkill. AWS Athena is a good place to start if you are just getting started on the cloud and want to test the waters at low cost and minimal effort. Athena however quickly runs into challenges with regards to limits, concurrency, transparency and consistent performance. You can find more details here. Costs will increase significantly as the scanned data volume grows. 

At Ahana, many of our customers are previous Athena and/or Redshift users that saw challenges around price performance (Redshift) and concurrency/deployment control (Athena). Keep in mind that Athena and Redshift Spectrum provide the same $5 terabyte scanned cost while Ahana is priced purely at instance hours. The power of Presto, ease of setup and management, price-performance, and dedicated compute resources. 
You can learn more about how Ahana compares to Amazon Athena here:

Ahana PAYGO Pricing

Ahana Cloud is easy to use, fully-integrated, and cloud native. Only pay for what you use with a pay-as-you-go model and no upfront costs.

Redshift Data Warehouse Architecture Explained

Amazon Redshift is a cloud data warehouse offered as a managed service by AWS. Learn more about what it is and how it differs from traditional data warehouses.

AWS Redshift Query Limits

At its heart, Redshift is an Amazon petabyte-scale data warehouse product that is based on PostgreSQL version 8.0.2.