Closed ianjevans closed 1 year ago
I'm working on repro'ing this ...
I got the above error message when using this main.tfvars file with the examples/workflows/cockroach_dedicated_cluster/main.tf file. Terraform provider version 0.3.2, Terraform version 1.3.4
cluster_name = "soper-version-test"
cloud_provider = "AWS"
cloud_provider_regions = [
"us-east-1",
"us-east-2",
"us-west-2"
]
machine_type = "m5.large"
I can't repro Ian's issue. When I use "v22.1", it sends the correct API request with that version string. I am seeing the same behavior as Mark, though, which suggests that the cluster might be auto-updating before marking itself as ready, or something like that. Digging in now.
Now it's working fine. Very strange.
Closing as a "can't reproduce."
I actually ended up hitting the "inconsistent results" error while trying to run acceptance tests. Turns out there's a version selection feature flag that's disabled by default. I'd expect a helpful error pointing the user toward their account rep to enable the flag if there's a version in the request, so I'll put something in JIRA.
I got an "unsupported version" error when running
terraform apply
using the dedicated workflow example: https://github.com/cockroachdb/terraform-provider-cockroach/blob/main/examples/workflows/cockroach_dedicated_cluster/main.tfI used the following
terraform.tfvars
file:The example
main.tf
specifies the default version as v22.1. Changing it to v22.2 allowed the plan to successfully apply.