Interpolated (count.index) route_table route re-create every run

[zantharax]

Hey there, I am seeing an issue while using an interpolated "gateway_id" for a route table route block. The configuration is below, but every time terraform runs a plan, it fails to recognize the existing route block, instead clearing the current route block and re-creating the route with the same information.

Terraform Version

v0.11.2

Affected Resource(s)

• route_table

Terraform Configuration Files

resource "aws_route_table" "route_table" {
  count = "3"
  vpc_id = "${aws_vpc.default.id}"

  route = {
    cidr_block = "0.0.0.0/0"
    gateway_id = "${element(aws_nat_gateway.natgw.*.id,count.index)}"
  }
}

resource "aws_nat_gateway" "natgw" {
  count = "3"
  allocation_id = "${element(aws_eip.eip.*.id,count.index)}"
  subnet_id = "${element(aws_subnet.subnets.*.id,count.index)}"
}

resource "aws_eip" "eip" {
  count = "3"
  vpc = true
}

resource "aws_subnet" "subnets" {
  count      = "3"
  cidr_block = "10.0.${count.index+200}.0/24"
  vpc_id     = "${aws_vpc.default.id}"

  availability_zone = "${element(data.aws_availability_zones.available.names,count.index)}"

}

data "aws_availability_zones" "available" {
}

Expected Behavior

Terraform plan should have interpreted the interpolated route as identical and not changed the route table and associated routes

Actual Behavior

Terraform clears and re-creates the route with interpolated gateway data.
E.G.

      route.3532503076.cidr_block:                "0.0.0.0/0" => ""
      route.3532503076.egress_only_gateway_id:    "" => ""
      route.3532503076.gateway_id:                "" => ""
      route.3532503076.instance_id:               "" => ""
      route.3532503076.ipv6_cidr_block:           "" => ""
      route.3532503076.nat_gateway_id:            "nat-####" => ""
      route.3532503076.network_interface_id:      "" => ""
      route.3532503076.vpc_peering_connection_id: "" => ""
      route.3761834389.cidr_block:                "" => "0.0.0.0/0"
      route.3761834389.egress_only_gateway_id:    "" => ""
      route.3761834389.gateway_id:                "" => "nat-####"
      route.3761834389.instance_id:               "" => ""
      route.3761834389.ipv6_cidr_block:           "" => ""
      route.3761834389.nat_gateway_id:            "" => ""
      route.3761834389.network_interface_id:      "" => ""
      route.3761834389.vpc_peering_connection_id: "" => ""

Steps to Reproduce

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

1. terraform plan/apply

References

Are there any other GitHub issues (open or closed) or Pull Requests that should be linked here? For example:

• Seems similar to resource/aws_instance: vpc_security_group_ids always showing update #1993 and other associated issues, but not sure if root cause is the same.
• May be a long term issue documented here: plan is in a loop changing my "main" route table #26

[zantharax]

Figured out the issue - AWS and the provider both allow for NAT gateways to be used in the gateway_id field. However, when aws provisions the resource, it is changed to the nat_gateway field, causing a diff in the state file.

Modifying the config to

resource "aws_route_table" "route_table" {
  count = "3"
  vpc_id = "${aws_vpc.default.id}"

  route = {
    cidr_block = "0.0.0.0/0"
    nat_gateway_id = "${element(aws_nat_gateway.natgw.*.id,count.index)}"
  }
}

Solved the issue

[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 feel this issue should be reopened, we encourage creating a new issue linking back to this one for added context. Thanks!