Terraform: Terraform Crash: empty structure "access_logs."

[jvasallo]

Hi there,

Terraform Version

v0.8.5

Affected Resource(s)

Not sure currently, We made no changes, just upgraded to latest TF. It appears to be related to walking aws_route53_record and trying to compare for access_logs? It could also be related to aws_elasticache_cluster as that is a dependency.

Terraform Configuration Files

#######
# Route53 CNAME for ElasticCache
#######
resource "aws_route53_record" "memcached_endpoint" {
  zone_id = "${var.private_dns_zone}"
  name = "memcached.application.date.env.company.com"
  type = "CNAME"
  ttl = "5"
  records = ["${element("${split(":","${aws_elasticache_cluster.memcached.configuration_endpoint}")}",0)}"]
}

#########################
#            ElastiCache               #
#########################
resource "aws_elasticache_cluster" "memcached" {
    cluster_id = "memcached"
    engine = "memcached"
    engine_version = "1.4.24"
    node_type = "cache.r3.large"
    num_cache_nodes = 2
    parameter_group_name = "default.memcached1.4"
    port = 12345
    subnet_group_name = "${aws_elasticache_subnet_group.cache_subnet.name}"
    security_group_ids = ["${aws_security_group.app_b.id}", "${aws_security_group.office_seattle.id}"]
}

Debug Output

See Crash.log (let me know if you need more info)

Panic Output

https://gist.github.com/jvasallo/cac90ce45fb4e7e6639d01744d900273

Expected Behavior

Terraform should apply successfully.

Actual Behavior

Terraform appears to panic post upgrade to v0.8.5.

Steps to Reproduce

1. terraform apply

Important Factoids

N/A

References

N/A

[jvasallo]

To help narrow down the issue, downgrading to Terraform v0.8.4 seems to have fixed the panic. I don't have time currently to dig into the issue more, but if this is not resolved later, I might take a stab at it. Sorry :(

[jvasallo]

After troubleshooting this further, it was not memcached related at all. The issue was we have an ELB we created through TF, an engineer manually added access_logs in the AWS Console to the ELB, when we run TF it was panicing about the diffs. I removed it from the ELB manually and things started working as expected.

I expect Terraform to complain about this, but it should toss a panic. Is that correct? I will test this issue in an environment that isn't production soon :joy:

[grubernaut]

Hi @jvasallo, thanks for the issue!

This has actually been fixed in #11321, and should be released in the next minor version of Terraform (v0.8.6). Closing the issue for now, but happy to discuss further! Thanks!

[hashibot[bot]]

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.