Apply fails when hpa metric type was changed in the cluster

[shysank]

Terraform Version, Provider Version and Kubernetes Version

Terraform version: 1.7.5
Kubernetes provider version: 2.31.0
Kubernetes version: 1.28

Affected Resource(s)

• kuberenetes_manifest

Terraform Configuration Files

resource "kubernetes_manifest" "hpa" {
  manifest        = yamldecode(file("path_to_hpa"))
  field_manager {
    force_conflicts = true
  }
}

Debug Output

Panic Output

Steps to Reproduce

1. Add a kuberetnes hpa manifest with cpu metric as a kuberetes_manifest resource
   Eg.

apiVersion: autoscaling/v2
kind: HorizontalPodAutoscaler
...
spec:
  scaleTargetRef:
    apiVersion: apps/v1
    kind: Deployment
  metrics:
    - type: Resource
      resource:
        name: cpu
        target:
          averageUtilization: 80
          type: Utilization 

3. Update the hpa directly in the cluster using kubectl edit to something like (note: target:)

spec:
  scaleTargetRef:
    apiVersion: apps/v1
    kind: Deployment
  metrics:
    - type: Resource
      resource:
        name: cpu
        target:
          averageValue: 100m
          type: AverageValue 

4. terraform apply

Expected Behavior

It should override the hpa with the configured spec in kuberenetes_manifest resource

Actual Behavior

Apply failed with the following error

╷
│ Error: Provider produced inconsistent result after apply
│ 
│ When applying changes to
│ module.service.kubernetes_manifest.service_components["autoscale.yaml"],
│ provider "provider[\"registry.terraform.io/hashicorp/kubernetes\"]"
│ produced an unexpected new value:
│ .object.spec.metrics[1].resource.target.type: was
│ cty.StringVal("Utilization"), but now cty.StringVal("AverageValue").
│ 
│ This is a bug in the provider, which should be reported in the provider's
│ own issue tracker

Important Factoids

References

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment

[yashrajdighe]

│ Error: Provider produced inconsistent result after apply
│ 
│ When applying changes to
│ kubernetes_manifest.deployment_kafka_strimzi_cluster_operator, provider
│ "provider[\"registry.terraform.io/hashicorp/kubernetes\"]" produced an
│ unexpected new value:
│ .object.spec.template.spec.containers[0].resources.limits["cpu"]: was
│ cty.StringVal("1000m"), but now cty.StringVal("1").
│ 
│ This is a bug in the provider, which should be reported in the provider's
│ own issue tracker.

Faced a similar issue.

[shysank]

Adding computed_fields resolved the original issue but a similar issue happened if the CPU target utilization was changed.

Steps to Reproduce

1. Repeat the original issue with computed_fields = ["metadata.annotations", "metadata.labels", "spec.metrics"]
2. Update spec.metrics[0].target. averageUtilization = 70
3. terraform plan
4. terraform apply

Expected Behavior
It should override the hpa with the configured spec in kuberenetes_manifest resource

Actual Behavior
Apply failed with the following error

╷
│ Error: Provider produced inconsistent result after apply
│ 
│ When applying changes to
│ module.service.kubernetes_manifest.service_components["autoscale.yaml"],
│ provider "provider[\"registry.terraform.io/hashicorp/kubernetes\"]"
│ produced an unexpected new value:
│ .object.metadata.annotations["analysis.scaleops.io/original-metrics"]: was
│ cty.StringVal("[{\"type\":\"Resource\",\"resource\":{\"name\":\"cpu\",\"target\":{\"type\":\"Utilization\",\"averageUtilization\":55}}},{\"type\":\"Pods\",\"pods\":{\"metric\":{\"name\":\"istio_requests_pod_per_second\"},\"target\":{\"type\":\"AverageValue\",\"averageValue\":\"300\"}}}]"),
│ but now
│ cty.StringVal("[{\"type\":\"Resource\",\"resource\":{\"name\":\"cpu\",\"target\":{\"type\":\"Utilization\",\"averageUtilization\":58}}},{\"type\":\"Pods\",\"pods\":{\"metric\":{\"name\":\"istio_requests_pod_per_second\"},\"target\":{\"type\":\"AverageValue\",\"averageValue\":\"300\"}}}]").
│ 
│ This is a bug in the provider, which should be reported in the provider's
│ own issue tracker.
╵

Note: The changes get applied. Running terraform plan again shows no infrastructure changes.

@alexsomesan Whenever you get a chance, it'd be great if you can share some thoughts on this. Thanks!