-
Notifications
You must be signed in to change notification settings - Fork 3
Open
Description
In order to test correctness and robustness of our framework implementation under failure, we need to able to simulate the node failure and communication failure. For now we can start with simulate the node failure. To do this, we simply randomly abort the goroutine (that represent the task).
The main change we need is to make sure that after a task failed (corresponding goroutine exits), a new goroutine will start to take its place.
@fengjingchao, can you start to work on this?
Metadata
Metadata
Assignees
Labels
No labels