wait-for-cluster.sh infinite loop when multiple clusters meet the filter name

[rwong2888]

terraform -v
Terraform v0.13.5

• provider registry.terraform.io/hashicorp/external v2.0.0
• provider registry.terraform.io/hashicorp/google v3.46.0
• provider registry.terraform.io/hashicorp/kubernetes v1.13.3
• provider registry.terraform.io/hashicorp/null v3.0.0
• provider registry.terraform.io/hashicorp/random v3.0.0

gcloud -v
Google Cloud SDK 316.0.0
alpha 2020.10.23
beta 2020.10.23
bq 2.0.62
core 2020.10.23
gsutil 4.54
kubectl 1.15.11

environment
abc-xyz already exists and created successfully via terraform
abc is new cluster that is stuck in terraform apply

gcloud container clusters list --project=REDACTED --filter=name:"abc"
NAME                   LOCATION  MASTER_VERSION   MASTER_IP     MACHINE_TYPE   NODE_VERSION     NUM_NODES  STATUS
abc             us-east4  1.16.13-gke.401  REDACTED  n1-standard-2  1.16.13-gke.401  3          RUNNING
abc-xyz  us-east4  1.16.13-gke.401  REDACTED  n1-standard-2  1.16.13-gke.401  4          RUNNING

behaviour:

module.gke.module.gcloud_wait_for_cluster.null_resource.run_command[0]: Provisioning with 'local-exec'...
module.gke.module.gcloud_wait_for_cluster.null_resource.run_destroy_command[0]: Creation complete after 0s [id=4940799199719502166]
module.gke.module.gcloud_wait_for_cluster.null_resource.run_command[0] (local-exec): Executing: ["/bin/sh" "-c" "PATH=/google-cloud-sdk/bin:$PATH\n.terraform/modules/gke/modules/private-cluster/scripts/wait-for-cluster.sh REDACTED REDACTED\n"]
module.gke.module.gcloud_wait_for_cluster.null_resource.run_command[0] (local-exec): Waiting for cluster REDACTED in project REDACTED to reconcile...
module.gke.module.gcloud_wait_for_cluster.null_resource.run_command[0] (local-exec): .
module.gke.module.gcloud_wait_for_cluster.null_resource.run_command[0] (local-exec): .
module.gke.module.gcloud_wait_for_cluster.null_resource.run_command[0]: Still creating... [10s elapsed]
module.gke.module.gcloud_wait_for_cluster.null_resource.run_command[0] (local-exec): .
module.gke.module.gcloud_wait_for_cluster.null_resource.run_command[0] (local-exec): .
module.gke.module.gcloud_wait_for_cluster.null_resource.run_command[0]: Still creating... [20s elapsed]
module.gke.module.gcloud_wait_for_cluster.null_resource.run_command[0] (local-exec): .
module.gke.module.gcloud_wait_for_cluster.null_resource.run_command[0]: Still creating... [30s elapsed]
module.gke.module.gcloud_wait_for_cluster.null_resource.run_command[0] (local-exec): .
module.gke.module.gcloud_wait_for_cluster.null_resource.run_command[0] (local-exec): .
module.gke.module.gcloud_wait_for_cluster.null_resource.run_command[0]: Still creating... [40s elapsed]
module.gke.module.gcloud_wait_for_cluster.null_resource.run_command[0] (local-exec): .
module.gke.module.gcloud_wait_for_cluster.null_resource.run_command[0] (local-exec): .
module.gke.module.gcloud_wait_for_cluster.null_resource.run_command[0]: Still creating... [50s elapsed]
module.gke.module.gcloud_wait_for_cluster.null_resource.run_command[0] (local-exec): .
...

expected behaviour:
expected to return after cluster is in running status

replication:
In wait-for-cluster.sh

current_status=$(gcloud container clusters list --project="$PROJECT" --filter=name:"$CLUSTER_NAME" --format="value(status)")

current_status="RUNNING RUNNING" instead of "RUNNING"

In this case the filter is returning multiple clusters instead of the cluster exactly named 'abc'

workaround:
Manually deleting cluster, firewall rules, and remote state. Passing skip_provisioners=true as per documentation and 413 to skip RUNNING check and wait for cluster RUNNING manually

[bharathkkb]

Thanks for reporting @rwong2888
Perhaps we can filter by endpoint which will be unique for a given cluster.

[bharathkkb]

Looks like we have been using key:pattern as opposed to key=value. I believe an equality match on name AND location should prevent this behavior.