newrelic_workflow | Provider tries to recreate workflow and crashes.

[AdrienKuhn]

Terraform Version

Terraform v1.2.7
on darwin_amd64
+ provider registry.terraform.io/newrelic/newrelic v3.2.0

Affected Resource(s)

• newrelic_notification_channel
• newrelic_workflow

Terraform Configuration

resource "newrelic_notification_channel" "test_channel" {
  account_id = "(supposed to be optional but is actually required, see https://github.com/newrelic/terraform-provider-newrelic/issues/1993)"
  name = "Test Channel"
  type  = "PAGERDUTY_SERVICE_INTEGRATION"
  destination_id = "(pagerduty_destination_id)"
  product  = "IINT"

  property {
    key = "summary"
    value = "{{ annotations.title.[0] }}"
    label = "PagerDuty Summary"
  }

}

resource "newrelic_workflow" "test_workflow" {
  destinations_enabled = false
  enrichments_enabled = false
  muting_rules_handling = "NOTIFY_ALL_ISSUES"
  name = "Test Workflow"
  workflow_enabled  = true

  issues_filter {
    name = "policyName"
    type = "FILTER"

    predicates {
      attribute = "accumulations.policyName"
      operator  = "CONTAINS"
      values = ["something"]
    }
  }

  destination_configuration {
    channel_id = newrelic_notification_channel.test_channel.id
  }
}

Actual Behavior

• First apply is working as expected
• Subsequent plan is trying to recreate the workflow

Terraform will perform the following actions:

  # module.newrelic_alayamarket_workflow.newrelic_workflow.test_workflow must be replaced
-/+ resource "newrelic_workflow" "test_workflow" {
      - account_id            = 123456 -> null # forces replacement
      ~ destinations_enabled  = true -> false
      ~ enrichments_enabled   = true -> false

     [...]

Plan: 1 to add, 0 to change, 1 to destroy.

• An apply will delete the workflow (which will also delete the channel on NewRelic side, which is unexpected) and the apply will fail

module.newrelic_alayamarket_workflow.newrelic_workflow.test_workflow: Destroying... [id=78f97f58-b884-4dae-9481-1a0a6a74b903]
module.newrelic_alayamarket_workflow.newrelic_workflow.test_workflow: Destruction complete after 3s
module.newrelic_alayamarket_workflow.newrelic_workflow.test_workflow: Creating...
╷
│ Error: INVALID_PARAMETER: A channel with id f35f0b02-8ba2-45a0-8f38-5e60b58e1f31 was not found
│
│   with module.newrelic_alayamarket_workflow.newrelic_workflow.test_workflow,
│   on modules/alert-workflow/main.tf line 52, in resource "newrelic_workflow" "test_workflow":
│   52: resource "newrelic_workflow" "test_workflow" {

• Subsequent applies make the provider crash

Stack trace from the terraform-provider-newrelic_v3.2.0 plugin:

panic: runtime error: index out of range [0] with length 0

goroutine 866 [running]:
github.com/newrelic/terraform-provider-newrelic/v2/newrelic.resourceNewRelicNotificationChannelRead({0x1c0a530, 0xc00005ac60}, 0xc000972500, {0x193fee0?, 0xc00090d1d0})
	/home/runner/work/terraform-provider-newrelic/terraform-provider-newrelic/newrelic/resource_newrelic_notifications_channel.go:132 +0x44e
github.com/hashicorp/terraform-plugin-sdk/v2/helper/schema.(*Resource).read(0xc0004dd500, {0x1c0a568, 0xc000868d20}, 0xd?, {0x193fee0, 0xc00090d1d0})
	/home/runner/go/pkg/mod/github.com/hashicorp/terraform-plugin-sdk/[email protected]/helper/schema/resource.go:724 +0x12e
github.com/hashicorp/terraform-plugin-sdk/v2/helper/schema.(*Resource).RefreshWithoutUpgrade(0xc0004dd500, {0x1c0a568, 0xc000868d20}, 0xc000a96ea0, {0x193fee0, 0xc00090d1d0})
	/home/runner/go/pkg/mod/github.com/hashicorp/terraform-plugin-sdk/[email protected]/helper/schema/resource.go:1015 +0x585
github.com/hashicorp/terraform-plugin-sdk/v2/helper/schema.(*GRPCProviderServer).ReadResource(0xc0001119c8, {0x1c0a568?, 0xc000868c00?}, 0xc0002a08c0)
	/home/runner/go/pkg/mod/github.com/hashicorp/terraform-plugin-sdk/[email protected]/helper/schema/grpc_provider.go:613 +0x497
github.com/hashicorp/terraform-plugin-go/tfprotov5/tf5server.(*server).ReadResource(0xc00046d2c0, {0x1c0a568?, 0xc000868600?}, 0xc00005a600)
	/home/runner/go/pkg/mod/github.com/hashicorp/[email protected]/tfprotov5/tf5server/server.go:747 +0x41e
github.com/hashicorp/terraform-plugin-go/tfprotov5/internal/tfplugin5._Provider_ReadResource_Handler({0x1a39e40?, 0xc00046d2c0}, {0x1c0a568, 0xc000868600}, 0xc0002703f0, 0x0)
	/home/runner/go/pkg/mod/github.com/hashicorp/[email protected]/tfprotov5/internal/tfplugin5/tfplugin5_grpc.pb.go:349 +0x170
google.golang.org/grpc.(*Server).processUnaryRPC(0xc00049e000, {0x1c0de70, 0xc000582340}, 0xc0008b6120, 0xc0004e7b60, 0x21d24f0, 0x0)
	/home/runner/go/pkg/mod/google.golang.org/[email protected]/server.go:1295 +0xb0b
google.golang.org/grpc.(*Server).handleStream(0xc00049e000, {0x1c0de70, 0xc000582340}, 0xc0008b6120, 0x0)
	/home/runner/go/pkg/mod/google.golang.org/[email protected]/server.go:1636 +0xa1b
google.golang.org/grpc.(*Server).serveStreams.func1.2()
	/home/runner/go/pkg/mod/google.golang.org/[email protected]/server.go:932 +0x98
created by google.golang.org/grpc.(*Server).serveStreams.func1
	/home/runner/go/pkg/mod/google.golang.org/[email protected]/server.go:930 +0x28a

Error: The terraform-provider-newrelic_v3.2.0 plugin crashed!

• The only way out is to manipulate the terraform state to delete the module reference.

Expected Behavior

• Plan should not try to recreate the workflow since there are no changes
• Deleting the workflow should not delete the channel since those are 2 distinct resources in the Terraform configuration.

Steps to Reproduce

Please list the steps required to reproduce the issue, for example:

1. Add the above configuration
2. terraform apply
3. terraform plan (this should show that it will recreate the workflow)
4. terraform apply (this will delete the workflow and the channel)
5. terraform apply (this will make the provider crash)

References

• May be related to https://github.com/newrelic/terraform-provider-newrelic/issues/1974, but in the case above there are no changes in the configuration.

[NSSPKrishna]

Hi @AdrienKuhn thanks for raising the issue. We will prioritise it.

[NSSPKrishna]

Hi @AdrienKuhn please try with the latest release V3.2.1

[arhill05]

I can confirm this still occurs on V3.2.1.

I can work around it by either performing a terraform destroy before the apply, or deleting the .tfstate file... but that causes other issues obviously.

[NSSPKrishna]

Ok, we will prioritise this issue.

[thoriqtoped]

I am having the same issue with version 3.4.1

[IliaShurygin]

Hi! Thank you for the detailed report! I have reproduced the issue and we are going to work on fixing it.

As a temporary workaround, you could specify account id on the workflow resources explicitly. This is obviously not ideal, but should at least unblock you until the problem is fixed.

[IliaShurygin]

The problem with account_id is fixed in v3.5.0, you should now be able to create (was already possible before) and update (the previously broken part) workflows without specifying account_id in workflow resource explicitly.