terraform-provider-aws 1.2 fails with "netrpc: connect: no such file or directory" on plan/apply when using with terraform 0.10

[endemics]

Hi,

We started seeing issues today with our CD pipeline today where the "terraform plan" and "terraform apply" commands started to fail with "netrpc: connect: no such file or directory".

Fixing the provider version to 1.1 using version = "= 1.1" fixed the issue, as well as using terraform 0.10.1 and above, so I believe it's related to the 1.2 release yesterday and similar to hashicorp/terraform#15756

Should the requirements in Readme be updated to indicate that version 1.2 and beyond need terraform 0.10.1 and above instead of 0.10.x?

[aaomoware]

i second that.

i'm seeing similiar issue:

➜  linux_amd64 git:(master) ✗ tf version          
Terraform v0.10.0

Your version of Terraform is out of date! The latest version
is 0.10.8. You can update by downloading from www.terraform.io
➜  linux_amd64 git:(master) ✗ ll                 
total 67M
-rwxr-xr-x 1 aaomoware aaomoware 159 Nov  2 10:38 lock.json
-rwxr-xr-x 1 aaomoware aaomoware 55M Nov  2 10:38 terraform-provider-aws_v1.2.0_x4
-rwxr-xr-x 1 aaomoware aaomoware 12M Nov  2 10:38 terraform-provider-consul_v1.0.0_x4
➜  linux_amd64 git:(master) ✗ 

tf plan fails with:

* module.env.provider.aws: dial unix /tmp/plugin117931723|netrpc: connect: no such file or directory

[aaomoware]

@endemics

tf upgrade resolved this issue for me

➜  env git:(master) ✗ tf version
Terraform v0.10.8

➜  env git:(master) ✗ ll .terraform/plugins/linux_amd64/
total 67M
-rwxr-xr-x 1 aaomoware aaomoware 159 Nov  2 10:52 lock.json
-rwxr-xr-x 1 aaomoware aaomoware 55M Nov  2 10:52 terraform-provider-aws_v1.2.0_x4
-rwxr-xr-x 1 aaomoware aaomoware 12M Nov  2 10:52 terraform-provider-consul_v1.0.0_x4
➜  env git:(master) ✗ 

[Yuxael]

Either update terraform to 0.10.8 or specifically set

provider aws {
    version = "1.1"
}

in all configurations because provider version 1.2 is not backward compatible with terraform <v0.10.8
👎

[bflad]

Given that this issue is related to very old provider/core versions from shortly after the provider split, I'm going to close this issue as most people should be on 0.10.8 or 0.11.x at this point. If someone does feel inclined to update documentation regarding this for some reason, we can certainly take a look.

[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!