v1.1.0

chore: bump security group module version @andrewhertog (#210)

what

• The security group module dependency has been upgraded to include a major fix to how it manages rules

why

• Using the old version made changes to security group rules hard to deal with.

references

https://github.com/cloudposse/terraform-aws-security-group/releases
https://sweetops.slack.com/archives/CB6GHNLG0/p1701898649784559

related

• Closes #192

v1.0.0

New variables: create_parameter_group and parameter_group_name @y3ti (#208)

what

• Add redis family suffix to parameter group name
• Add new variables: create_parameter_group and parameter_group_name

why

This module doesn't currently support major version upgrades of Redis (eg, 6.x to. 7.x) because:

• Parameter groups are major-version specific, so when a user changes var.family from redis6 to redis7, Terraform needs to create a new parameter group. Without create_before_destroy, Terraform tries to first destroy the old Param group which fails because it's currently in use
• Parameter groups must have unique names across families. When Terraform tries to create a new param group for redis7, it fails because it tries to do so using the same name as the old param group.

I have decided to add "redis cluster family" as a suffix. AWS follows a similar convention for default parameter groups, using names such as:

• default.redis6.x
• default.redis7

Since using . is not possible, I have opted to use - instead.

To prevent any breaking changes, I have introduced a new variable called parameter_group_name. By setting this variable to the current parameter group name, you can prevent any terraform configuration drift.

We can also reuse existing parameter groups.

If we want to use the default parameter group created by AWS (default.redis7)

create_parameter_group = false
engine                 = "redis7"

If we want to use any other existing parameter group:

create_parameter_group = false
parameter_group_name   = "existing-parameter-group-name"

references

Resolves #178

I see that other people tried to solve this problem before (see references), but the pull requests were not merged yet:

• #159
• #189
• #201
• #208

v0.53.0

Allow Module to Disable or Enable Cluster Mode for Existing Deployment @milldr (#181)

what

• added compact for when resource dne

why

• Currently, if the module is deployed with cluster mode enabled, it cannot be disabled without first destroying and redeploying

references

• Closes #180

v0.52.0

Support AWS Provider V5 @max-lobur (#203)

what

Support AWS Provider V5
Linter fixes

why

Maintenance

references

https://github.com/hashicorp/terraform-provider-aws/releases/tag/v5.0.0

v0.51.1

Sync github @max-lobur (#196)

Rebuild github dir from the template

🚀 Enhancements

feat: Add support for aws provider 5.0 @MaxymVlasov (#198)

why

That field was deprecated in 4.0 and was removed in 5.0

Note:

make init
make github/init
make readme

Already ran and committed

v0.51.0

• No changes

v0.50.0

Groundwork new workflows @max-lobur (#193)

Fix lint/format before workflows rollout

v0.49.0

Fix breaking change introduced in 0.46.0 - `user_group_ids` conflicts… @MaxymVlasov (#184)

… with auth_token

why

• If you specify only auth_token - Terraform plan will fail because it interprets user_group_ids = [] as specified value

Before 0.46.0 I can do next:

module "redis" {
  source  = "cloudposse/elasticache-redis/aws"
  version = "0.45.0"

  auth_token         = join("", random_password.auth_token[*].result)
}

After upgrade I need to do next, which I don't like:

module "redis" {
  source  = "cloudposse/elasticache-redis/aws"
  version = "0.48.0"

  # Confilicting parameters. Could be set only one of them
  auth_token         = join("", random_password.auth_token[*].result)
  user_group_ids     = null
}

So I just return previous logic

module "redis" {
  source = "git::https://github.com/MaxymVlasov/terraform-aws-elasticache-redis?ref=83f0142d97123157482d05c62eaaeecbacfb71e1"

  auth_token         = join("", random_password.auth_token[*].result)
}

v0.48.0

feat: Add `data_tiering_enabled` @Cheezmeister (#175)

what

• Introduce var.data_tiering_enabled and plumb it through to the aws provider

why

• @teikametrics needs to set data_tiering_enabled true to provision our desired node type

  Error: error creating ElastiCache Replication Group (bidder-redis-cluster-production): InvalidParameterCombination: When     using the cache.r6gd.xlarge node type, you must enable data tiering.
  status code: 400, request id: REDACTED
  

• This flag data_tiering_enabled doesn't currently exist in the terraform-aws-elasticache-redis module

references

• https://registry.terraform.io/providers/hashicorp/aws/latest/docs/resources/elasticache_replication_group#data_tiering_enabled

v0.47.0

Add `auto_minor_version_upgrade` parameter @LieneJansone (#183)

what

• Add auto_minor_version_upgrade parameter

why

• To allow disabling the Auto upgrade minor versions option

references

• closes #182