You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently, results data is kept by the scheduler until the client it's associated disconnects.
Ideally, we could discard the task result once received by the client:
In [1]: importmathIn [2]: fromscalerimportClientIn [3]: client=Client("tcp://127.0.0.1:60338")
# ScalerTop shows 1 scheduler object (the serializer)In [4]: future=client.submit(math.cos, 0)
# ScalerTop shows 4 scheduler objects (the serializer, the function, the function argument and the return value).In [5]: future.result()
Out[5]: 1.0# ScalerTop still shows 4 objects, while the scheduler could clear the task's result data as it reached the client's future. In [6]: client.disconnect()
# ScalerTop shows 0 scheduler object
The text was updated successfully, but these errors were encountered:
Note that the client cannot always send a delete request when receiving the result object, as this result object might be required by another dependent graph task.
Currently, results data is kept by the scheduler until the client it's associated disconnects.
Ideally, we could discard the task result once received by the client:
The text was updated successfully, but these errors were encountered: