Function: set_stream_failure_mode

Applicability

Product	Supported
Atlas 350 Accelerator Card	√
Atlas A3 training product/Atlas A3 inference product	√
Atlas A2 training product/Atlas A2 inference product	√
Atlas training product	√
Atlas inference product	√
Atlas 200I/500 A2 inference product	√

Description

When multiple tasks are delivered to a stream, you can use this API to specify the task scheduling mode to determine whether to continue the next task after a task fails.

Prototype

C Prototype

aclError aclrtSetStreamFailureMode(aclrtStream stream, uint64_t mode)

Python Function

ret = acl.rt.set_stream_failure_mode(stream, mode)

Parameters

Parameter	Description
stream	Int, pointer address of the stream to be operated. The support for the default stream (that is, this parameter is set to 0) varies according to product models. Atlas 350 Accelerator Card: not supported Atlas A3 training product/Atlas A3 inference product: supported Atlas A2 training product/Atlas A2 inference product: supported Atlas 200I/500 A2 inference product: not supported Atlas inference product: not supported Atlas training product: not supported
mode	Int, the task scheduling mode to determine whether to continue the next task after a task fails, when multiple tasks are delivered to a stream. The options are as follows: ACL_CONTINUE_ON_FAILURE = 0: default value. If a task fails, the system continues to execute the next task. ACL_STOP_ON_FAILURE = 1: After a task fails, the subsequent tasks are stopped.

Parameter

Description

stream

Int, pointer address of the stream to be operated.

The support for the default stream (that is, this parameter is set to 0) varies according to product models.

Atlas 350 Accelerator Card: not supported

Atlas A3 training product/Atlas A3 inference product: supported

Atlas A2 training product/Atlas A2 inference product: supported

Atlas 200I/500 A2 inference product: not supported

Atlas inference product: not supported

Atlas training product: not supported

mode

Int, the task scheduling mode to determine whether to continue the next task after a task fails, when multiple tasks are delivered to a stream.

The options are as follows:

ACL_CONTINUE_ON_FAILURE = 0: default value. If a task fails, the system continues to execute the next task.
ACL_STOP_ON_FAILURE = 1: After a task fails, the subsequent tasks are stopped.

Return Value

Return Value	Description
ret	Int, error code. 0 on success; else, failure.

Restrictions

This API can be called only once for a specified stream to set the task scheduling mode.
If the stop-on-failure mode is set for a stream, the stop-on-failure mode also applies to other streams in the context of the stream. This restriction applies to the following product models:
Atlas A3 training product/Atlas A3 inference product

Atlas A2 training product/Atlas A2 inference product

Atlas 200I/500 A2 inference product

Atlas inference product

Atlas training product

Parent topic: Streams Management