[Bug]: aws_elasticache_serverless_cache recreates if engine is changed from valkey 7 to redis 7

[sukumaar]

Terraform and AWS Provider Version

Terraform Core Version
v1.5.7

AWS Provider Version
v6.20.0

Affected Resource(s) or Data Source(s)

aws_elasticache_serverless_cache

Expected Behavior

Terraform should allow in-place valkey 7 to redis 7 engine update as AWS console allows same behaviour. It should not destroy and recreate resource.

Actual Behavior

Terraform is recreating elasticache serverless.

Terraform used the selected providers to generate the following execution plan. Resource actions are
indicated with the following symbols:
-/+ destroy and then create replacement

Terraform will perform the following actions:

  # aws_elasticache_serverless_cache.test_redis_valkey must be replaced
-/+ resource "aws_elasticache_serverless_cache" "test_redis_valkey" {
      ~ arn                      = "arn:aws:elasticache:us-west-2:123456789:serverlesscache:test-redis" -> (known after apply)
      ~ create_time              = "2025-11-21T16:36:57Z" -> (known after apply)
      ~ daily_snapshot_time      = "11:30" -> (known after apply)
      ~ description              = " " -> (known after apply)
      ~ endpoint                 = [
          - {
              - address = "test-redis-qwerty.serverless.usw2.cache.amazonaws.com"
              - port    = 6379
            },
        ] -> (known after apply)
      ~ engine                   = "valkey" -> "redis" # forces replacement
      ~ full_engine_version      = "7.2" -> (known after apply)
      ~ id                       = "test-redis" -> (known after apply)
        name                     = "test-redis"
      ~ reader_endpoint          = [
          - {
              - address = "test-redis-qwerty.serverless.usw2.cache.amazonaws.com"
              - port    = 6380
            },
        ] -> (known after apply)
      ~ snapshot_retention_limit = 0 -> (known after apply)
      ~ status                   = "available" -> (known after apply)
        # (5 unchanged attributes hidden)
    }

Plan: 1 to add, 0 to change, 1 to destroy.

Changes to Outputs:
  ~ elasticache_engine               = "valkey" -> "redis"
  ~ elasticache_primary_endpoint     = [
      - {
          - address = "test-redis-qwerty.serverless.usw2.cache.amazonaws.com"
          - port    = 6379
        },
    ] -> (known after apply)
  ~ elasticache_serverless_cache_id  = "test-redis" -> (known after apply)

Relevant Error/Panic Output

Sample Terraform Configuration

initial valkey 7

resource "aws_elasticache_serverless_cache" "test_redis_valkey" {
  name   = "test-redis"
  
  engine = "valkey"  
  
  major_engine_version = "7"

  subnet_ids = [
    aws_subnet.test_subnet_a.id,
    aws_subnet.test_subnet_b.id
  ]

  security_group_ids = [aws_security_group.elasticache_sg.id]
}

after redis 7

resource "aws_elasticache_serverless_cache" "test_redis_valkey" {
  name   = "test-redis"

  engine = "redis"
  
  major_engine_version = "7"

  subnet_ids = [
    aws_subnet.test_subnet_a.id,
    aws_subnet.test_subnet_b.id
  ]

  security_group_ids = [aws_security_group.elasticache_sg.id]
}

Steps to Reproduce

1. create aws_elasticache_serverless_cache with engine valkey 7
2. terraform plan & terraform apply
3. update engine to redis 7
4. terraform plan

Debug Logging

Click to expand log output

GenAI / LLM Assisted Development

n/a

Important Facts and References

No response

Would you like to implement a fix?

Yes

[github-actions]

Community Guidelines

This comment is added to every new Issue to provide quick reference to how the Terraform AWS Provider is maintained. Please review the information below, and thank you for contributing to the community that keeps the provider thriving! 🚀

Voting for Prioritization

• Please vote on this Issue by adding a 👍 reaction to the original post to help the community and maintainers prioritize it.
• Please see our prioritization guide for additional information on how the maintainers handle prioritization.
• Please do not leave +1 or other comments that do not add relevant new information or questions; they generate extra noise for others following the Issue and do not help prioritize the request.

Volunteering to Work on This Issue

• If you are interested in working on this issue, please leave a comment.
• If this would be your first contribution, please review the contribution guide.
• For new resources and data sources, use skaff to generate scaffolding with comments detailing common expectations.