Crash during destroy.

[clstokes]

During a destroy I experienced a crash. I was destroying 200 machines and the crash happened at the very end. Interestingly, when I re-ran the destroy it completed immediately with no machines leftover from when the crash occurred. I double-checked and indeed all machines had been destroyed.

Terraform Version

$ terraform version
Terraform v0.11.1
+ provider.triton v0.3.0

Affected Resource(s)

• triton_machine

Terraform Configuration Files

data "triton_image" "ubuntu" {
  name        = "ubuntu-16.04"
  type        = "lx-dataset"
  most_recent = true
}

resource "triton_machine" "bastion" {
  count = "${var.bastion_machine_count}"

  name    = "bastion-${count.index}"
  package = "g4-general-8G"
  image   = "${data.triton_image.ubuntu.id}"

  tags {
    role = "bastion"
  }

  cns {
    services = ["bastion"]
  }
}

Output

https://gist.github.com/clstokes/9bbac5320a687fcd3c23aa3056ae8608#file-_tf-log

Panic Output

https://gist.github.com/clstokes/9bbac5320a687fcd3c23aa3056ae8608#file-crash-log

Expected Behavior

Destroy should have completed successfully.

Actual Behavior

Terraform crashed.

Steps to Reproduce

Apply and then destroy. I don't think this will be easily reproducible though.

[clstokes]

Another crash - different part of code, but I think same general issue - crash on EOF.

https://gist.github.com/clstokes/35dcedbb3188edd06305a908ee06472e#file-_tf-log

[clstokes]

Oh... able to reproduce the one from my last comment relatively easier. Block the the Triton provider executable from accessing the network (using Tripmode or something similar) and you'll get the EOF.

[clstokes]

Still getting a similar error with Triton provider 0.4.0.

https://gist.github.com/clstokes/587ea4ddc5f75c12ef0670577d3dfb3a