Closed cgillum closed 2 months ago
Looks like the failing CI is new and not part of our normal CIs. I think we can safely ignore its results.
@cgillum: agreed. Surprised to see it running automatically on every PR, but that's fine since it's just tests.
This is to allow exception handling code to catch a more specific exception and differentiate between unexpected errors and expected errors.
This is largely non-breaking because
OrchestrationAlreadyExistsException
already derives fromInvalidOperationException
.Note that the
IOrchestrationServiceClient
interface already documents that theCreateTaskOrchestrationAsync
methods should be throwingOrchestrationAlreadyExistsException
, so this change fixes the AzureStorageOrchestrationService's implementation to follow the existing spec.