Closed kaibocai closed 11 months ago
Also giving a timeout for the purging instance call seems meaningless, even if it's timeout, eventually, the instance will be purged as all this happens at the storage backend.
A timeout accomplishes two things:
Depending on the implementation, a client-specified timeout can still cause the server-side operation to be canceled. For example, in .NET, this can be done by using the cancellation token provided in the gRPC server-side context to any downstream purge operations.
Let me know if I'm misunderstanding.
Make sense thank you!
I noticed that the
PurgeInstanceCriteria
class has atimeout
field https://github.com/microsoft/durabletask-java/blob/b3e2d3ab9dfc07495f830d90b00d7fac35009ea0/client/src/main/java/com/microsoft/durabletask/PurgeInstanceCriteria.java#L19 which is used to timeout thepurgeInstance
Grpc call.This
timeout
field is not found in durabletask-dotnet. Also giving a timeout for the purging instance call seems meaningless, even if it's timeout, eventually, the instance will be purged as all this happens at the storage backend.I think we should remove this timeout parameter. @cgillum, @DeepanshuA