Terraform0.12 with aws2.24 crashes during plan

[korjek]

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
• Please do not leave "+1" or "me too" comments, they generate extra noise for issue followers and do not help prioritize the request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment

Terraform Version

Terraform version 0.12.6
aws provider version: 2.24

Terraform Configuration Files

https://www.dropbox.com/s/gmgneeyhvyyqn6a/gen.tf?dl=0
https://www.dropbox.com/s/u3rczydswur1fd2/spot_fleet_az.tf?dl=0

Panic Output

https://gist.github.com/korjek/c91a42f752175f50c152c0e1d2db8d05

Expected Behavior

Terraform should produce plan without crashing

Actual Behavior

Terrafrom keeps crashing

Steps to Reproduce

1. terraform plan

Important Factoids

We don't have this issue with aws provider 2.16

[bflad]

Relevant portion of the crash log:

2019/08/19 15:44:47 Instance store backed AMIs do not provide a root device name - Use an EBS AMI
panic: Instance store backed AMIs do not provide a root device name - Use an EBS AMI

goroutine 225 [running]:
log.Panic(0xc000c99890, 0x1, 0x1)
 /opt/goenv/versions/1.12.6/src/log/log.go:333 +0xac
github.com/terraform-providers/terraform-provider-aws/aws.launchSpecsToSet(0xc000c8ca90, 0x19, 0x19, 0xc0000ae330, 0xc000b47190)
 /opt/teamcity-agent/work/5d79fe75d4460a2f/src/github.com/terraform-providers/terraform-provider-aws/aws/resource_aws_spot_fleet_request.go:979 +0x13d
github.com/terraform-providers/terraform-provider-aws/aws.resourceAwsSpotFleetRequestRead(0xc000230070, 0x4963be0, 0xc0005d0000, 0xc000230070, 0x0)
 /opt/teamcity-agent/work/5d79fe75d4460a2f/src/github.com/terraform-providers/terraform-provider-aws/aws/resource_aws_spot_fleet_request.go:969 +0x4d9

Terraform resource code should never call panic() (or in this case log.Panic()) and instead should return a normal error to the operator.

This is related to #9775 as they share the same underlying function call to fetchRootDeviceName().

Definition of Done

Presumably, this must be done before fixes from #9775 are implemented so the error returns the unexpected crash.

• Add covering test to aws/resource_aws_spot_fleet_request_test.go that triggers the current log.Panic()
• Change the launchSpecsToSet function signature to include returning an error, e.g. launchSpecsToSet(launchSpecs []*ec2.SpotFleetLaunchSpecification, conn *ec2.EC2) (*schema.Set, error)
• Switch implementation of return log.Panic(err) to return nil, err (and final return specSet to return specSet, nil)
• Appropriately bubble up any returned errors from calls to launchSpecsToSet
• Verify error is returned instead of panic

At this point, this crash issue can be closed out. The fixes from #9775 will fix the underlying implementation issue.

[ghost]

This has been released in version 2.29.0 of the Terraform AWS provider. Please see the Terraform documentation on provider versioning or reach out if you need any assistance upgrading.

For further feature requests or bug reports with this functionality, please create a new GitHub issue following the template for triage. Thanks!

[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!