# Outdated AWS Glue version for Spark ETL jobs Canonical: https://www.pointfive.co/efficiency-hub/inefficiencies/outdated-aws-glue-version-for-python-jobs-d2d59 Newer AWS Glue versions - such as Glue 5.0, with 5.1 (now the default for new jobs) and 6.0 also available - include significant performance... By: Nick Sudarikov Updated: 2026-09-28 [Cloud Efficiency Hub](https://www.pointfive.co/efficiency-hub) # Outdated AWS Glue Version for Spark ETL Jobs [Nick Sudarikov](https://www.linkedin.com/in/nikolaysudarikov/) Finance Director, Technology Operations @ InvestCloud, Inc. AWS service [AWS Glue](https://www.pointfive.co/efficiency-hub/cloud-services/aws-glue) Category [Compute](https://www.pointfive.co/efficiency-hub/service-category/compute) Reference CER-0249 Type Outdated Version ## Explanation Why the waste happens and who it affects. Newer AWS Glue versions - such as Glue 5.0, with 5.1 (now the default for new jobs) and 6.0 also available - include significant performance optimizations in the Spark runtime for Glue for Spark ETL jobs, whether written in PySpark or Scala. AWS reports that Glue 5.0 is 32% faster than Glue 4.0 and reduces costs by 22% on its benchmarks. These improvements often require few or no code changes, making version upgrades a simple and impactful optimization. When jobs remain on older runtimes such as Glue 3.0 or 4.0 (or Glue 2.0 and earlier, which reached end of life on April 1, 2026), they execute more slowly, consume more DPUs, and incur unnecessary cost. Jobs on Glue 4.0 or later can also use larger G.12X and G.16X workers and memory-optimized R workers, which can provide additional performance gain for some jobs. ## Billing model The pricing dimensions that drive this cost. AWS Glue is billed per second based on the number of Data Processing Units (DPUs) used. Faster runtime versions reduce job execution time, lowering total DPU consumption and overall cost. Older versions can extend runtime duration without providing any functional benefit. ## How to detect 4 checks to find it in your estate. - Review Glue job configurations to identify jobs still running on older Glue versions such as 3.0 or 4.0, or end-of-life 2.0 and earlier - Confirm the job is a Glue for Spark ETL job (PySpark or Scala), since the performance gains come from the Spark runtime - Assess job execution times that remain consistently longer than expected for similar workloads - Evaluate whether runtime upgrades have been incorporated into standard Glue job governance or deployment templates ## How to fix 4 ways to remove the waste. - Upgrade Glue for Spark jobs to the latest supported Glue runtime version to benefit from performance improvements - Incorporate runtime version management into IaC templates and deployment pipelines to ensure consistent modernization - Periodically review job runtimes to confirm alignment with current Glue versions and available optimizations - Validate job output after upgrading to confirm compatibility with updated library and Spark versions ## Documentation Vendor references for pricing and configuration. - [https://docs.aws.amazon.com/glue/latest/dg/release-notes.html docs.aws.amazon.com](https://docs.aws.amazon.com/glue/latest/dg/release-notes.html) - [Introducing AWS Glue 5.0 for Apache Spark aws.amazon.com](https://aws.amazon.com/blogs/big-data/introducing-aws-glue-5-0-for-apache-spark/) - [Migrating AWS Glue for Spark jobs to AWS Glue version 5.0 docs.aws.amazon.com](https://docs.aws.amazon.com/glue/latest/dg/migrating-version-50.html) ## Related inefficiencies [Browse the library](https://www.pointfive.co/efficiency-hub) - AWS Glue CER-0393 ### [Non-Urgent Glue Spark Jobs Not Using Flex Execution](https://www.pointfive.co/efficiency-hub/inefficiencies/non-urgent-glue-spark-jobs-not-using-flex-execution) AWS Glue Spark jobs run in the standard execution class unless configured otherwise, which provides fast startup on dedicated resources and is billed at the standard DPU-hour rate. Glue also offers a Flex execution class that runs jobs on... Compute - AWS Glue CER-0394 ### [Glue Spark Jobs With Fixed Worker Counts Instead of Auto Scaling](https://www.pointfive.co/efficiency-hub/inefficiencies/glue-spark-jobs-with-fixed-worker-counts-instead-of-auto-scaling) AWS Glue Spark jobs are commonly configured with a fixed NumberOfWorkers sized for the heaviest stage, the largest expected input or a generous guess. Without Auto Scaling, every allocated worker is billed for the whole run, including... Compute - AWS Lambda CER-0209 ### [Unnecessary Costs from Unused Lambda Versions with SnapStart](https://www.pointfive.co/efficiency-hub/inefficiencies/unnecessary-costs-from-unused-lambda-versions-with-snapstart) Many teams publish new Lambda versions frequently (e.g., through CI/CD pipelines) but do not clean up old ones. When SnapStart is enabled on Python or .NET functions, each of these versions retains a cached snapshot, generating ongoing... Compute --- Source: the public page above. Product screenshots and illustrative interfaces are examples, not live customer data.