Pop up `panic: can't use ElementIterator on null value` error
Opened this issue · 3 comments
Hi @jckuester , recently I got the following error while cleanup ecs resource. the weird thing is current region doesn't have any ecs cluster, but it still happens each time.
Type: aws_ecs_cluster
Found: 1
Id: arn:aws:ecs:ap-southeast-2:xxxxxxxxxxxx:cluster/default
---
• TOTAL NUMBER OF RESOURCES THAT WOULD BE DELETED: 1
• USER WILL NOT BE ASKED FOR CONFIRMATION (FORCE MODE)
• STARTING TO DELETE RESOURCES
panic: can't use ElementIterator on null value
goroutine 194 [running]:
github.com/zclconf/go-cty/cty.Value.ElementIterator(0x5f98800, 0xc00048b450, 0x0, 0x0, 0x7, 0x0)
/home/jan/go/pkg/mod/github.com/zclconf/go-cty@v1.4.0/cty/value_ops.go:1038 +0x11a
github.com/zclconf/go-cty/cty.Value.AsValueMap(0x5f98800, 0xc00048b450, 0x0, 0x0, 0x1)
/home/jan/go/pkg/mod/github.com/zclconf/go-cty@v1.4.0/cty/value_ops.go:1243 +0xfe
github.com/jckuester/terradozer/pkg/provider.enableForceDestroyAttributes(0x5f98800, 0xc00048b450, 0x0, 0x0, 0x203000, 0x203000, 0x203000, 0xf4a6622f9a8800fb)
/home/jan/go/pkg/mod/github.com/jckuester/terradozer@v0.1.2/pkg/provider/provider.go:236 +0xf7
github.com/jckuester/terradozer/pkg/provider.TerraformProvider.DestroyResource.func1(0x4c0da0)
/home/jan/go/pkg/mod/github.com/jckuester/terradozer@v0.1.2/pkg/provider/provider.go:200 +0x86
github.com/hashicorp/terraform/helper/resource.Retry.func1(0x0, 0x0, 0x0, 0x0, 0x0, 0x0)
/home/jan/go/pkg/mod/github.com/hashicorp/terraform@v0.12.25/helper/resource/wait.go:22 +0x5e
github.com/hashicorp/terraform/helper/resource.(*StateChangeConf).WaitForState.func1(0xc00926e720, 0xc0001d4540, 0xc000850600, 0xc007e240c0, 0xc0082242d0, 0xc0082242c8)
/home/jan/go/pkg/mod/github.com/hashicorp/terraform@v0.12.25/helper/resource/state.go:103 +0x298
created by github.com/hashicorp/terraform/helper/resource.(*StateChangeConf).WaitForState
/home/jan/go/pkg/mod/github.com/hashicorp/terraform@v0.12.25/helper/resource/state.go:80 +0x1bf
The awsweeper policy is following with latest version v.0.10.1
aws_ecs_cluster:
- tags:
NOT(Owner): .*
Could you help to take a look?
Hi @showerlee again,
thanks for another bug report. I guess know what the issue is and will fix it:
I assume that your ECS cluster is in state INACTIVE (which means it has already been deleted). Can you please check? AWSweeper should not list INACTIVE ECS clusters, but it currently does - I will fix that. I also will fix the panic. What happen there is that I assume that we won't see already deleted resources in the deletion stage (with nil state) and don't check for nil there.