AWS Glue Streaming Job Timeout Cannot Be Reset to Unlimited in Terraform – HashiCorp Help Center

Introduction

When managing AWS Glue Streaming Jobs through Terraform, users may configure a custom timeout for their jobs. However, historically, once a timeout value was set, reverting it back to the default unlimited (null) state was not supported. This resulted in Terraform being unable to clear the timeout configuration, leading to drift and configuration mismatches between Terraform state and AWS.

Problem

Users managing AWS Glue streaming jobs via Terraform observed the following behavior:

Creating a job without specifying a timeout results in:
- Terraform state showing timeout = 0
- AWS CLI showing None (unlimited)
After setting a timeout, for example timeout = 500, the value updates successfully in AWS.
Attempting to revert the timeout to unlimited fails:
- timeout = 0 → Terraform error:
  Error: expected timeout to be at least (1), got 0
- Removing the timeout argument or setting it to null shows no diff in plan, and AWS job does not update.
AWS CLI also prevented clearing the timeout:
- Setting Timeout = 0 → parameter validation error
- Sending an empty update request → InvalidInputException: Command should not be null

From the AWS Console, users could clear the timeout manually, but Terraform and AWS CLI lacked support for doing the same.

Prerequisites

AWS Glue Streaming ETL Job
Terraform AWS Provider (affected versions before 6.20.0)
Terraform CLI

Cause

In earlier versions of the AWS Provider ( Till 6.19.0 ) , the timeout argument for aws_glue_job enforced a validation rule requiring a minimum value of 1, making it impossible to set the value back to 0 to represent "unlimited".

Solutions:

The issue has been resolved in the AWS Provider version 6.20.0.

Fix Introduced:

timeout = 0 is now allowed for Apache Spark Streaming ETL jobs.
Setting timeout = 0 now correctly configures the Glue job to have no timeout (unlimited).
Terraform will properly update the Glue job and reflect the change in state and AWS.

Example Updated Configuration

resource "aws_glue_job" "streaming_job" {
  name              = "sample-streaming-job-repro"
  role_arn          = aws_iam_role.glue_job_role.arn
  glue_version      = "5.0"
  timeout           = 0 ## Allowed from provider version 6.20.0+
  number_of_workers = 2
  max_retries       = 0
  worker_type       = "G.1X"
  execution_class   = "STANDARD"

Additional Information

https://registry.terraform.io/providers/hashicorp/aws/6.20.0/docs/resources/glue_job#timeout-5