[Bug]: serverless elasticache major version upgrade recreates a new resource #41644
Description
Terraform Core Version
1.7.5
AWS Provider Version
5.76.0
Affected Resource(s)
aws_elasticache_serverless_cache
Expected Behavior
AWS allows for upgrading the major_engine_version attribute of an existing serverless cache via the AWS console or using the Elasticache API, see: https://docs.aws.amazon.com/AmazonElastiCache/latest/dg/VersionManagement.html#VersionManagement-serverless
Updating the major engine version from "7" to "8" should not trigger a destroy + recreation of an existing resource.
Actual Behavior
Modifying the major engine version from "7" to "8" on a serverless cluster resource does not attempt to do an in-place version upgrade that AWS supports, instead the provider wants to destroy the cluster and create a new one, see:
Relevant Error/Panic Output Snippet
# aws_elasticache_serverless_cache.dev-serverless-valkey must be replaced
-/+ resource "aws_elasticache_serverless_cache" "dev-serverless-valkey" {
~ full_engine_version = "7.2" -> (known after apply)
~ id = "dev-serverless-valkey" -> (known after apply)
~ major_engine_version = "7" -> "8" # forces replacement
<several other attributes indicating that they will be known after apply>Terraform Configuration Files
resource "aws_elasticache_serverless_cache" "dev-serverless-valkey" {
name = "dev-serverless-valkey"
description = "Valkey serverless cache"
engine = "valkey"
major_engine_version = "8"
subnet_ids = [, , ]
security_group_ids = []
user_group_id =
}
Steps to Reproduce
- create a aws_elasticache_serverless_cache resource with the major_engine_version attribute set to "7"
- update the major_engine_version value to "8"
- run a plan
- observe that the provider wants to replace the existing serverless cache with a completely new serverless cache, rather than doing the in-place upgrade
Debug Output
No response
Panic Output
No response
Important Factoids
No response
References
No response
Would you like to implement a fix?
None