"Unexpected EOF" and "connection is shut down" with google providers when using terraform 0.7.1

[sbadia]

Hi there,

I just switched to terraform 0.7.1, my infra is a mix of ±200 ressources on gcp + openstack (so multi provider).

Note: Everything work's like a charm with terraform 0.7.0

Terraform Version

• terraform 0.7.1

Affected Resource(s)

• google_*

Debug Output

• http://pub.sebian.fr/pub/crash-run.log.gpg
  • size:8417
  • hash:b56ac0776faa3f638bf919ba130509fb
  • crypted with 0x51852D87348FFC4C

Panic Output

• http://pub.sebian.fr/pub/crash.log.gpg
  • size:279154
  • hash:90280650f20ba041310e84f247cc1b9a
  • crypted with 0x51852D87348FFC4C

Expected Behavior

What should have happened?

Terraform should report or execute the plan for google_records

Actual Behavior

What actually happened?

Any google_record throws either a "Unexpected EOF" and "connection is shut down" error.

Note: I tried with -parallelism=1 as suggested in #8011 (comment) but no change…

Steps to Reproduce

1. terraform plan

Important Factoids

Are there anything atypical about your accounts that we should know? For example: Running in EC2 Classic? Custom version of OpenStack? Tight ACLs?

No

References

Maybe this one (but for cloudfare ressources)

• "Unexpected EOF" and "connection is shut down" with Cloudflare provider  #8011

Thanks in advance!

[vaygr]

Same issue here.

[jaygorrell]

Same here from Atlas with AWS. Worked great on 0.7.0.

[cblecker]

Hi @sbadia!
From what I can see "unexpected EOF" usually relates to an issue with your local system communicating with the provider's API. Also, what OS/arch are you running on your local system (uname -a)?

@vaygr reported a similar issue in #8345, but I wasn't able to duplicate it using Mac OS X 10.11.6 (Darwin x86_64). I can't see your debugs as they're encrypted and I'm not a member of the terraform team, but perhaps you can compare them to @vaygr's to see if they are similar.

[jen20]

Looks like this is potentially related to #8345, we'll try to ensure this is fixed and released in 0.7.2.

[sbadia]

hi, sorry for the delay!

@cblecker I use Debian GNU/Linux with a 4.6.0 kernel
Linux baloo 4.6.0-1-amd64 #1 SMP Debian 4.6.4-1 (2016-07-18) x86_64 GNU/Linux

the panic error seems indeed related panic: interface conversion: interface is nil, not string

Just re-tested with master Terraform v0.7.2-dev (a69c62af655184bbb164f3516929c830862c57ed) I still have the issue with a plan :-/

• crash.log terraform v0.7.2-dev

[rvrignaud]

Same issue described above by @sbadia on same terraform files have been reproduced on OS X 10.11

[sbadia]

The PR #8390 solve this issue. Thanks @jen20 !

❯ curl -s https://patch-diff.githubusercontent.com/raw/hashicorp/terraform/pull/8390.patch | git apply -
❯ make
[…]
❯ XC_OS=linux XC_ARCH=amd64 make bin
[…]
❯ terraform version
Terraform v0.7.2-dev (51d669be44a696c868faa8719253859fac312d4a+CHANGES)

Everything re-work fine :)

[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.