Possible leak of ELB in HCP cluster

    Whilst creating CI pipeline for terraform HCP modules and I seem to be getting a somewhat consistent issue where I'm not able to destroy my VPC due to dependencies

DependencyViolation: The subnet 'subnet-042834d44276dc194' has dependencies and cannot be deleted. status code: 400, request id: 5bb9dfd5-9b01-48dd-85da-5148e4c6b812

With help from QE (CI runs on their AWS account) we have determined the dependency to be an ELB

"Seems ELB is leak which is not deleted with cluster deletion. arn:aws:elasticloadbalancing:us-west-2:301721915996:loadbalancer/net/a0f5c134a3c3f4e85aec027f2e256f19/59b4b2533b3ad69c Need to research more. This is leak by ingress service."

Looking through the service logs this cluster (2allrsrpj9ouscic03fre11aedae9g2o - tf-ci-b2d8 - b0db2635-3a99-42b6-9779-77cb2af21481) doesn't seem to have been deleted improperly or through best effort

Is leaking ELB something that is already being tracked or some new issue? This test was initially conducted on production environment by the way, we are now moving to staging

Cluster is created with default version at the time (4.14.20)

4.14.20

    Fairly consistent, yesterday only 2 out of 8 (maybe more) runs didn't had this issue in production env

    1. Create AWS resources (account/operator roles, oidc config/provider and VPC)
    2. Create HCP cluster with IDP
    3. Delete HCP cluster and wait until it is removed from OCM
    4. Soon after removal call AWS API to delete resources, VPC errored on dependency
    

    VPC errored on dependency

    VPC can be deleted successfuly

Leaked LB info
https://files.slack.com/files-tmb/T030G10V24F-F06UEUE6R45-c1d5b9e843/image_720.png