Workflows AWS Connector Execution - Memory Limitations

Daniel_Kaufmann — Wed, 09 Jul 2025 13:19:48 GMT

Summary

When leveraging Dynatrace Workflows to automate your AWS infrastructure using AWS Connector, you can target a rich set of services, such as Amazon Web Services S3 (Simple Storage Service), AWS EC2 (Elastic Compute Cloud) or AWS SSM (Systems Manager).

If you want to list resources of any kind, such as Buckets from Amazon Web Services S3, Instances from AWS EC2 or list documents from AWS SSM, you may encounter issues. This article will walk you through a memory limitation problem that might occur and explain how to resolve it.

Problem

When running an AWS Connector Action with the goal to list resources, the execution might crash, causing the execution to abort with an error such as:

Error running action: Execution crashed - OOM - error ref 76e23678-80f5-4845-b82b-aed80dfeb53f

In the case of an execution crash when listing resources like in the error message shown above, it could be that the queried result is not limited in size, resulting in an OOM (Out-Of-Memory) error. This is caused by the result size being larger than the memory limitations in the Dynatrace platform.

Resolution

Each AWS Connector Action used for listing resources offers a pagination token or maximum result field in their Action Widget that can be used for achieving this.

The naming of these fields can vary per action depending on what Amazon Web Services defined in their official documentation, e.g. MaxItems, MaxResults, MaxBuckets, MaxParts etc. for maximum result or NextToken, ContinuationToken, Marker etc. for the pagination token. Please refer to the official documentation to find the corresponding field names for the action you are troubleshooting.

Example: EC2 Describe Instances Crashes with Out-Of-Memory Error

For example, executing the ec2-describe-instances action with the goal to list all AWS EC2 (Elastic Compute Cloud) instances, it is possible limit the result by utilizing the MaxResults or NextToken (also see the official documentation) fields:

Alternatively, it’s also possible to specify InstanceIds or use Filters to narrow down the results.

What's next