Upgrading from 0.39.0 to 0.44.0 fails with lots of errors

mission-lblaney commented 2 years ago

Describe the Bug

When upgrading from one minor version to another the terraform plan fails with about 50+ errors.

Expected Behavior

When upgrading minor versions I expect that planning will not fail because the changes should be backwards compatible.

Steps to Reproduce

I have 3 elastic cache clusters deployed with 0.39.0, upgrading them to 0.44.0 caused about 53 errors each. I did not change the values to the modules arguments and I did not change terraform version or aws provider version.

Screenshots

Environment (please complete the following information):

This is Terraform enterprise running terraform version 1.0.9. I'm not sure about the AWS provider version. I dont see one specified, so I'm assuming it's the latest, but I will ask a co-worker (I'm new to Terraform Enterprise).

mission-lblaney commented 2 years ago

I also ran this on the latest terraform version, 1.2.4 and it still failed to plan for the same errors.

mission-lblaney commented 2 years ago

I finally found this https://github.com/cloudposse/terraform-aws-elasticache-redis/blob/master/docs/migration-notes-0.41.0.md while going through every single release notes.

I'm guessing I will need to upgrade to 0.41.0 before I can upgrade to later versions. Will try that and report back.

mission-lblaney commented 2 years ago

0.41.0 also fails, same errors

mission-lblaney commented 2 years ago

I have narrowed this down to us still using allowed_cidr_blocks. I'm not sure when it became unusable but it seems like if we use the newer additional_security_group_rules our troubles will go away?

I'm not sure what you will do with this issue but I think its still valid. 44 straight minor releases but backwards compatibility has been broken since at least 0.41.0

cloudposse / terraform-aws-elasticache-redis