Closed dkropachev closed 7 months ago
I am not sure this PR closes the issue.
We are trading runtime panics (caused by invalid api responses) for runtime errors (introduced here) - and both of them have the same effect - user is not able to use the provider.
Would it be possible to dig deeper what causes those invalid responses that terraform provider is receiving?
@rjeczalik , GCP VPC Peering is giving out 0 in region_id
all the time, that is why p.RegionByID(vpcPeering.RegionID)
returns nil and we have panic here
Scenario described in https://github.com/scylladb/siren/issues/10469: Clone this repo https://github.com/ricardoborenstein/1m-ops-demo/tree/gcp
Endup successfully after this patch:
scylladbcloud_vpc_peering.scylladbcloud: Destroying... [id=scylla-cloud-vpc-297-peering-15e834ad67b5fc07f5a2]
scylladbcloud_vpc_peering.scylladbcloud: Destruction complete after 1s
scylladbcloud_cluster.scylladbcloud: Destroying... [id=297]
scylladbcloud_cluster.scylladbcloud: Destruction complete after 1s
Destroy complete! Resources: 2 destroyed.
Thanks for explaining it to me @dkropachev!
Closing https://github.com/scylladb/siren/issues/10469