[BUG] ovh_cloud_project_kube_nodepool default timeouts too low

[royolsen]

Describe the bug

create nodepool will often fail due to timeout.

Terraform Version

N/A

OVH Terraform Provider Version

ovh/ovh v0.42.0

Affected Resource(s)

• ovh_cloud_project_kube_nodepool

Terraform Configuration Files

resource "ovh_cloud_project_kube_nodepool" "node_pool" {
   service_name  = "${var.service_name}"
   kube_id       = ovh_cloud_project_kube.my_kube_cluster.id
   name          = "nodepool"
   flavor_name   = "d2-8"
   desired_nodes = 3
   max_nodes     = 3
   min_nodes     = 3
}

Expected Behavior

Terraform will wait until nodepool is crated, then continue execution.

Actual Behavior

Terraform throws a timeout error if creation takes more than 20 minutes. Pool is eventually created but terraform execution is halted on the error, leaving behind a terraform state out of sync with the created resources.

Steps to Reproduce

1. terraform apply

Workaround

Can be worked around by declaring specific timeouts in the resource specification:

resource "ovh_cloud_project_kube_nodepool" "node_pool" {

  timeouts {
    create = "1h"
    update = "1h"
    delete = "1h"
  }

   service_name  = "${var.service_name}"
   kube_id       = ovh_cloud_project_kube.my_kube_cluster.id
   name          = "nodepool"
   flavor_name   = "d2-8"
   desired_nodes = 3
   max_nodes     = 3
   min_nodes     = 3
}

Additional context

OVH will very often not spin up nodes within the default 20 minute timeout. While default timeout can be overridden, I suggest increasing the default timeout to reflect a realistic range of creation times. This will remove the need for adding an override, and reduce friction for new users.