When the new watch tasks are registered SDK queries homes for failed state. If a network goes down (during test scenario for instance) the task will cause unhandled rejection, or the runtime will hang waiting for intervals. I believe, @ltchang2019 has a good plan to finish his work on it, but still, in order to make the sdk-failed test case pass, I proposed shutting down those intervals in the SDK.
When the new watch tasks are registered SDK queries homes for failed state. If a network goes down (during test scenario for instance) the task will cause unhandled rejection, or the runtime will hang waiting for intervals. I believe, @ltchang2019 has a good plan to finish his work on it, but still, in order to make the
sdk-failed
test case pass, I proposed shutting down those intervals in the SDK.