AWS RDS - Restoring from snapshot doesn't set DBname

[kxseven]

In a Nutshell

• Creating an RDS instance by restoring from an existing Snapshot where the DBname will change results in a new RDS instance getting created with the DBname from the Snapshot.
• Subsequent 'terraform apply' actions detect the incorrect DBname and force a new (RDS) resource, Re-creating the problem all over again

Example

• An RDS snapshot is created myapp-sample-db from an existing RDS instance
• A new RDS instance is requested using the myapp-sample-db Snapshot identifier but with a different DBname MYAPPDEV

Terraform Version

Terraform v0.7.7, Terraform v0.7.9

Affected Resource(s)

• aws_db_instance

Terraform Configuration Files

resource "aws_db_instance" "development" {

    identifier                 = "myppdevdb"
    allocated_storage          = 500
    storage_type               = "gp2"
    engine                     = "oracle-ee"
    engine_version             = "12.1.0.2.v5"
    instance_class             = "db.r3.2xlarge"

    apply_immediately          = "true"

    password                   = "<redacted>"
    username                   = "<redacted>"
    name                       = "MYAPPDEV"

    publicly_accessible        = "false"
    db_subnet_group_name       = "<redacted>"
    parameter_group_name       = "<redacted>"
    option_group_name          = "<redacted>"
    copy_tags_to_snapshot      = "true"

    multi_az                   = "true"
    storage_encrypted          = "true"
    kms_key_id                 = "${var.kms_key_id}"
    vpc_security_group_ids     = [ "<redacted>" ]

    backup_retention_period    = 7
    backup_window              = "23:00-23:45"
    auto_minor_version_upgrade = "true"

    snapshot_identifier        = "myapp-sample-db"

    skip_final_snapshot        = "true"
    final_snapshot_identifier  = "devdb-final-snapshot"

}

Expected Behavior

• A new RDS instance should be created as a RestoreDBInstanceFromDBSnapshot applying the DBname change if specified in the HCL

Actual Behavior

• The RDS instance gets created from the Snapshot but with the DBname stored in the Snapshot
• Subsequent 'terraform apply' actions detect the incorrect DBname and force a new (RDS) resource, Re-creating the problem all over again

Steps to Reproduce

1. terraform apply (RDS instance gets created)
2. terraform apply (DBName is detected as incorrect, forces new resource)
3. terraform apply (ad infinitum)

References

• I believe the issue is due to the DBname not getting passed to the 'restore' method/function invocation (https://github.com/hashicorp/terraform/blob/master/builtin/providers/aws/resource_aws_db_instance.go#L378)
• Possibly also related to AWS RDS - Restoring from snapshot doesn't set username & password #8604
• Works as expected when using the AWS Console

[kxseven]

Is this a dupe of #4671 ?

If the root-cause is the AWS API (and not Terraform) should there not at least be a note in the TF docs to indicate this?

[ghost]

I'm going to lock this issue because it has been closed for 30 days ⏳. This helps our maintainers find and focus on the active issues.

If you have found a problem that seems similar to this, please open a new issue and complete the issue template so we can capture all the details necessary to investigate further.