[DOCS] Handling use case where cloud service broker is cycled or updated while performing terraform apply

[Bryan Kelly]

Interested to know what happens to the terraform apply process when the cloud service broker is cf pushed, restaged or terminated during a repave of the VM in which its hosted on? Will the terraform apply be retried or will the cloud service broker stop accepting requests and drain the current ones before terminating gracefully?

[Bret Mogilefsky]

I found that when a broker instance was restarted due to hitting a memory limit during a `terraform apply`, the API continues to return the "operation pending" state forever, and the terraform operation is never re-attempted. This should definitely be addressed.

…

On Tue, Feb 2, 2021 at 2:14 PM Bryan Kelly ***@***.***> wrote: Interested to know what happens to the terraform apply process when the cloud service broker is cf pushed, restaged or terminated during a repave of the VM in which its hosted on? Will the terraform apply be retried or will the cloud service broker stop accepting requests and drain the current ones before terminating gracefully? — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#170>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAALVEUZEIHNIUY274ZVVALS5B2KVANCNFSM4W7WZPNA> .

[Bryan Kelly]

yes just noticed this same behavior when attempting to delete a service instance and i believe the broker got turned over in the middle resulting in the same outcome.

[Ernie Billing]

There is currently no support in the broker to handle interrupted operations.

[Bret Mogilefsky]

Hey folks, any word on plans to address this? We think we ran into it again.

[Marcela Campo]

We are not planning to address this in the close future. Closing this for the moment.