redshift cluster restore from snapshot misconfigures cluster for parameters number_of_nodes and kms_key

[imdhruva]

Facing a similar issue here. I am trying to create a redshift cluster from snapshot. the restore from snapshot option enables me to change the existing 7 node ds2 based cluster to a 2 node ra3.4xl cluster. However, terraform seems to ignores all the configs and

• spins up a new 7 node ra3.4xl cluster.
• enable kms encryption on the cluster

Terraform Version

Terraform v0.12.12
+ provider.aws v2.60.0

Affected Resource(s)

• aws_redshift_cluster

Terraform Configuration Files

resource "aws_redshift_cluster" "main" {
  cluster_identifier                  = "xxx-redshift-cluster"
  database_name                       = "dev" # explicitly defining this to default to avoid confusion
  node_type                           = "ra3.4xlarge"
  cluster_subnet_group_name           = "sdata-redshift"
  preferred_maintenance_window        = "sun:06:25-sun:06:55"
  automated_snapshot_retention_period = 1
  number_of_nodes                     = 2
  publicly_accessible                 = false
  snapshot_identifier                 = "rs:xxxxxxxxxxxxxx"
  snapshot_cluster_identifier         = "xxxxxxxxxxxxxx"
  final_snapshot_identifier           = "xxx-redshift-cluster-snapshot-final"
  vpc_security_group_ids              = ["sg-beaa1bc7"]
  tags = {
    vpc   = "xxx"
    role  = "redshift"
    class = "cluster"
  }
}

Debug Output

https://gist.github.com/imdhruva/91911af2f501fe319337d2253481edca

Expected Behavior

aws_redshift_cluster with 2 nodes and no kms enabled

Actual Behavior

aws_redshift_cluster with 7 nodes and kms enabled

Steps to Reproduce

terraform init
terraform apply

References

#11367

[andy68-bris]

I can confirm I see the same behavior going the other way.
Current 7 node ds2 based cluster.
Terraform config to create a new 20 node ra3.4xl based on ds2 snapshot actually creates a 7 node ra3.4xl cluster.
Subsequent plan then detects a change from 7 nodes to 20. However I think this would trigger a classic resize rather than an elastic one.

[dpacaud]

This just bit us quite hard by spawning a 29 node cluster instead of 5 :(

[rednuht]

We are also experiencing that tags are not set when restoring from a snapshot. A second apply afterwards will correct this but when restoring from a snapshot of our main cluster's size it takes many hours... In our case that means that the cost allocation is completely wrong for a huge cluster for many hours, totally unacceptable!

[github-actions]

This functionality has been released in v4.9.0 of the Terraform AWS Provider. Please see the Terraform documentation on provider versioning or reach out if you need any assistance upgrading.

For further feature requests or bug reports with this functionality, please create a new GitHub issue following the template. Thank you!

[github-actions]

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.