cloudflare_logpush_job plan always suggests changes

[joshuamsager]

Confirmation

• [x] This is a bug with an existing resource and is not a feature request or enhancement. Feature requests should be submitted with Cloudflare Support or your account team.
• [x] I have searched the issue tracker and my issue isn't already found.
• [x] I have replicated my issue using the latest version of the provider and it is still present.

Terraform and Cloudflare provider version

Terraform v1.10.3 cloudflare/cloudflare v5.3.0

Affected resource(s)

• cloudflare_logpush_job

Terraform configuration files

resource "cloudflare_logpush_job" "kitchensink" {
  name             = "HTTP-requests-kitchensink"
  zone_id          = var.zone_id
  kind             = var.kind
  dataset          = "http_requests"
  destination_conf = "https://${var.destination}?header_Authorization=Basic%20${var.auth_token}"

  # Used only for "Core" logging (kind is null)
  max_upload_interval_seconds = 30
  max_upload_records          = 1000

  output_options = {
    field_names      = local.fields
    timestamp_format = "unixnano"
    sample_rate      = var.sample
    output_type      = "ndjson"
    record_prefix    = "{"
    record_suffix    = "}\n"
    field_delimiter  = ","
  }

  enabled = var.enabled
}

Link to debug output

Shared in Cloudflare Support ticket

Panic output

No response

Expected output

Plan after apply should not propose changes

Actual output

module.observe_apps.cloudflare_logpush_job.kitchensink will be updated in-place

~ resource "cloudflare_logpush_job" "kitchensink" { + error_message = (known after apply) id = 677037 - kind = "" -> null ~ last_complete = "2025-04-15T14:03:29Z" -> (known after apply) + last_error = (known after apply) name = "HTTP-requests-kitchensink" ~ output_options = { + cve_2021_44228 = false # (7 unchanged attributes hidden) } # (7 unchanged attributes hidden) }

Steps to reproduce

1. Provision a logpush job without a kind attribute
2. Configure output_options
3. Apply configuration
4. terraform plan

Additional factoids

No response

References

No response

[jhutchings1]

I've opened an internal issue to track the investigation on this.

[anordby]

I have this issue also. Please keep us posted @jhutchings1

[joshuamsager]

Still broken in v5.5.0

[dduportal]

Still an issue with v5.6.0 for info (ref. https://github.com/jenkins-infra/cloudflare/pull/60)

[jhutchings1]

The team believed this was fixed in v5.5.0. I have reactivated the internal issue and pinged engineering to take another look.

[dduportal]

The team believed this was fixed in v5.5.0. I have reactivated the internal issue and pinged engineering to take another look.

Thanks! If you need any tests we are happy to try (since most of the Jenkins infrastructure code is public and Cloudflare sponsors us).

[kmbzdyk]

This is still happening in v5.7.0

[mojanjz]

this issue https://github.com/cloudflare/terraform-provider-cloudflare/issues/5766 now is blocking us from reproducing this issue on v5.7.0

[kmbzdyk]

Update: I found out that the workaround for this is to remove the resource from UI and let terraform recreate it. I do use deprecated 'logpull_options' property instead of 'output_options' tho

[anordby]

This is still a problem in provider version 5.7.1. I am using output_options and don't want to use deprecated logpull_options.

Tried tainting and removing logpush in Cloudflare UI to recreate logpush resources. That works ok but the freshly created logpush resources still gives a recurring change

[MG-Nix]

Same issue with v5.7.1 Terraform v1.12.2

[rupalims]

This has been release in 5.8.2

[jamie-oconnell]

Just tested this with 5.8.2 and there is still drift with error_message, last_complete, last_error, output_options.sample_rate

[anordby]

This was solved in 5.8.2 for me

[jhutchings1]

Just tested this with 5.8.2 and there is still drift with error_message, last_complete, last_error, output_options.sample_rate

@jamie-oconnell do you mind opening a new issue with your repro? We're not able to reproduce what you're mentioning, so this will help us get to the bottom of it. Thanks!

[egarbi]

@jamie-oconnell, It happened to me that I had explicitly added output_options.sample_rate = 0 in an attempt to solve the perpetual issue. After upgrading the version and removing this line, the issue is now gone. Check if this is the case for you