Can't setup Route53 Failover DNS records - InvalidInput: Invalid Request

[ninjaMog]

Having trouble setting up failover DNS records via terraform, if I drop the health_check_id and failover fields, it all works fine.

I can setup the required DNS record manually with the same values. In fact if I do, when I run a plan terraform recognises it exists and says there's no changes to make.

Unfortunately no useful output from terraform, either when running dist version (Terraform v0.6.6), or a dev compiled version with debug!

dns.tf:

resource "aws_route53_record" "www" {
   zone_id = "ZNO4H1L7085BF"
   name = "test"
   health_check_id = "0e1a2063-19f4-4f27-8a39-10d8664515f7"
   failover = "PRIMARY"
   type = "CNAME"
   ttl = "60"
   set_identifier = "1"
   records = ["project-main-r6ubntg45w.elasticbeanstalk.com"]
}

Output from plan:

+ aws_route53_record.www
    failover:           "" => "PRIMARY"
    fqdn:               "" => "<computed>"
    health_check_id:    "" => "0e1a2063-19f4-4f27-8a39-10d8664515f7"
    name:               "" => "test"
    records.#:          "" => "1"
    records.2776629804: "" => "project-main-r6ubntg45w.elasticbeanstalk.com"
    set_identifier:     "" => "1"
    ttl:                "" => "60"
    type:               "" => "CNAME"
    zone_id:            "" => "ZNO4H1L7085BF"


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

terraform apply

aws_route53_record.www: Creating...
  failover:           "" => "PRIMARY"
  fqdn:               "" => "<computed>"
  health_check_id:    "" => "0e1a2063-19f4-4f27-8a39-10d8664515f7"
  name:               "" => "test"
  records.#:          "" => "1"
  records.2776629804: "" => "project-main-r6ubntg45w.elasticbeanstalk.com"
  set_identifier:     "" => "1"
  ttl:                "" => "60"
  type:               "" => "CNAME"
  zone_id:            "" => "ZNO4H1L7085BF"
Error applying plan:

1 error(s) occurred:

* aws_route53_record.www: InvalidInput: Invalid request
    status code: 400, request id: <requestid>

[ninjaMog]

Issue seems to be resolved in

Terraform v0.6.7-dev (0cbf79d+CHANGES)

(compiled from branch :- f-aws-elastic-beanstalk)

[teewhey]

I'm using version 0.6.9. I'm hitting this error when trying to destroy these resources.

[IanGordonOne]

Using terraform_0.6.11_windows_amd64.

main.tf:

resource "aws_route53_health_check" "green-mba" {
  fqdn = "montereybayaquarium.org"
  port = "80"
  type = "HTTP"
  resource_path = "/"
  failure_threshold = "3"
  request_interval = "30"

  tags = {
    Name = "MBA CMS health check"
  }
}

resource "aws_route53_record" "green-mba" {
  zone_id = "Z245YJSJRAN83I"
  name = "green-${var.environment}.${var.platform}.montereybayaquarium.org"
  type = "A"
  failover = "PRIMARY"
  health_check_id = "${aws_route53_health_check.green-mba.id}"

  alias {
    name = "${aws_elb.green-mba.dns_name}"
    zone_id = "${aws_elb.green-mba.zone_id}"
    evaluate_target_health = true
  }
}

resource "aws_route53_record" "green-mba-failover" {
  zone_id = "Z245YJSJRAN83I"
  name = "green-${var.environment}.${var.platform}.montereybayaquarium.org"
  type = "A"
  ttl = "300"
  records = ["${var.ip_of_failover_service}"]
  failover="SECONDARY"
}

Plan:

+ aws_route53_record.green-mba
    alias.#:                                 "" => "1"
    alias.1741685063.evaluate_target_health: "" => "1"
    alias.1741685063.name:                   "" => "green-production-cms-mba-979165135.us-west-1.elb.amazonaws.com"
    alias.1741685063.zone_id:                "" => "Z1M58G0W56PQJA"
    failover:                                "" => "PRIMARY"
    fqdn:                                    "" => "<computed>"
    health_check_id:                         "" => "072cab47-e03a-4e20-b2aa-187c0325a380"
    name:                                    "" => "green-production.cms.montereybayaquarium.org"
    type:                                    "" => "A"
    weight:                                  "" => "-1"
    zone_id:                                 "" => "Z245YJSJRAN83I"

+ aws_route53_record.green-mba-failover
    failover:          "" => "SECONDARY"
    fqdn:              "" => "<computed>"
    name:              "" => "green-production.cms.montereybayaquarium.org"
    records.#:         "" => "1"
    records.182719784: "" => "168.62.29.203"
    ttl:               "" => "300"
    type:              "" => "A"
    weight:            "" => "-1"
    zone_id:           "" => "Z245YJSJRAN83I"

Apply:

* aws_route53_record.green-mba-failover: InvalidInput: Invalid request
        status code: 400, request id: 88b608cc-d5c6-11e5-9a4c-8bab79036e13

[rajinders]

I am also trying to configure failover and running into the same issue described above by gordonlanJ
I am on terraform_0.6.12_windows_amd64

[rajinders]

It turns out that if I set the value of property set_identifier my route53 failover records are created.
Terraform documentation mentioned that this property is only required for weighted records so I was not setting it. When I was manually configuring the route53 recordset for failover I noticed that portal was automatically setting the value off ID. I tried doing the same and was pleased to see that record was successfully created.

[catsby]

Hello all –

I apologize for the trouble and silence here.

The crux of most of theses problems falls into to parts:

• the need for set_identifier is not properly documented for using failover records
• errors in deleting/modifying Route 53 Records from Terraform version 0.6.9 and earlier on newer versions

I've opened a pull request to address the documentation here:

• provider/aws: Document set_identifer requirement for Route53 weighted and failover types #5591

For issues with deleting the weighted records, that was fixed in #4892 , upgrading to the latest version of Terraform should address this.

Sorry again for the trouble! Let us know if you need anything else.

[ghost]

I'm going to lock this issue because it has been closed for 30 days ⏳. This helps our maintainers find and focus on the active issues.

If you have found a problem that seems similar to this, please open a new issue and complete the issue template so we can capture all the details necessary to investigate further.