I'm not entirely sure how to re-run terraform tasks from here, but one did fail for staging.

I think maybe that failed due to something else happening in Terraform at the same time; @cecilia-donnelly do you see anything about the change to staging that you would expect to cause an error?

You know, @slifty, this is a real problem that only shows up on staging! @Fenn-CS 's automated test script creates a subnet in Permanent's AWS that conflicts with the subnet search for staging, which is why this doesn't work. We ran into this a few months ago, too.

Last time I deleted the conflicting subnet since Fon wasn't running his tests then and had deleted the relevant instances. We could (1) do that again and bring it back up with the script afterward, (2) do that and also change the script so it uses a non-conflicting subnet, or (3) make the search in infrastructure a bit smarter so the subnets don't conflict anymore. Any of those is fine, though (3) requires a bit more careful testing.

@cecilia-donnelly I vote for 2 -- it sounds like a bug in the test scripts that should be resolved.

@Fenn-CS is this something you are able to knock out so we can get this merged?

@slifty Yeah early next week, but if it's urgent I can check it out this weekend.

@slifty So I started by doing 1) above which is deleting the subnets, using terraform destroy from the same environment that created the subnet.

I have looked into how to make it "smart" and found the cidrsubnet which I have played with locally and realized the addressed calculated are still pretty predictable and is not all that "smart".

So,

1. Either I have not figured out things correctly
2. It's as it seems, which means I don't really need cidrsubnet and can just change the conflicting IP to another one different from the existing one.

I want us to test this, by you repeating the action, so I can see if the cidrsubnet works as it should when the first address is unavailable before concluding what goes into the PR for the rclone-iac

For now, there is not conflict and the deployment can continue.

Further context

Here is the problematic section:

https://github.com/OpenTechStrategies/permanent-rclone-iac/blob/057d71a8a515d0fb541d75155336a602725d1761/main.tf#L40-L47

resource "aws_vpc" "mainvpc" {
  cidr_block = "10.0.0.0/16"
  enable_dns_support   = true
  enable_dns_hostnames = true
}

resource "aws_subnet" "public" {
  vpc_id = aws_vpc.mainvpc.id
  cidr_block = "10.0.0.0/24"
  availability_zone = "us-west-2a"
  tags = {
    Name = "Default subnet for us-west-2a"
  }
}

Here is what it should look like with cidrsubnet

resource "aws_vpc" "mainvpc" {
  cidr_block = "10.0.0.0/16"
  enable_dns_support   = true
  enable_dns_hostnames = true
}

resource "aws_subnet" "public" {
  vpc_id = aws_vpc.mainvpc.id
  cidr_block = cidrsubnet(aws_vpc.mainvpc.cidr_block, 8, 1) // The last paramameter as "0" leads to the exact same conflicting address
  availability_zone = "us-west-2a"
  tags = {
    Name = "Default subnet for us-west-2a"
  }
}