ray.experimental.state.common.TaskState#

class ray.experimental.state.common.TaskState(task_id: str, attempt_number: int, name: str, state: typing_extensions.Literal[NIL, PENDING_ARGS_AVAIL, PENDING_NODE_ASSIGNMENT, PENDING_OBJ_STORE_MEM_AVAIL, PENDING_ARGS_FETCH, SUBMITTED_TO_WORKER, RUNNING, RUNNING_IN_RAY_GET, RUNNING_IN_RAY_WAIT, FINISHED, FAILED], job_id: str, node_id: str, actor_id: str, type: typing_extensions.Literal[NORMAL_TASK, ACTOR_CREATION_TASK, ACTOR_TASK, DRIVER_TASK], func_or_class_name: str, language: str, required_resources: dict, runtime_env_info: str, parent_task_id: str, placement_group_id: str, worker_id: str, events: List[dict], profiling_data: List[dict], creation_time_ms: Optional[int], start_time_ms: Optional[int], end_time_ms: Optional[int])[source]#

Bases: ray.experimental.state.common.StateSchema

Task State

Below columns can be used for the --filter option.

actor_id

name

language

func_or_class_name

worker_id

type

parent_task_id

node_id

task_id

attempt_number

placement_group_id

state

job_id

Below columns are available only when get API is used,

--detail is specified through CLI, or detail=True is given to Python APIs.

language

events

creation_time_ms

start_time_ms

end_time_ms

placement_group_id

required_resources

profiling_data

runtime_env_info

task_id: str#

The id of the task.

attempt_number: int#

The attempt (retry) number of the task.

name: str#

The name of the task if it is given by the name argument.

state: typing_extensions.Literal[NIL, PENDING_ARGS_AVAIL, PENDING_NODE_ASSIGNMENT, PENDING_OBJ_STORE_MEM_AVAIL, PENDING_ARGS_FETCH, SUBMITTED_TO_WORKER, RUNNING, RUNNING_IN_RAY_GET, RUNNING_IN_RAY_WAIT, FINISHED, FAILED]#

The state of the task.

Refer to src/ray/protobuf/common.proto for a detailed explanation of the state breakdowns and typical state transition flow.

job_id: str#

The job id of this task.

node_id: str#

Id of the node that runs the task. If the task is retried, it could contain the node id of the previous executed task. If empty, it means the task hasn’t been scheduled yet.

actor_id: str#

The actor id that’s associated with this task. It is empty if there’s no relevant actors.

type: typing_extensions.Literal[NORMAL_TASK, ACTOR_CREATION_TASK, ACTOR_TASK, DRIVER_TASK]#

The type of the task.

  • NORMAL_TASK: Tasks created by func.remote()`

  • ACTOR_CREATION_TASK: Actors created by class.remote()

  • ACTOR_TASK: Actor tasks submitted by actor.method.remote()

  • DRIVER_TASK: Driver (A script that calls ray.init).

func_or_class_name: str#

The name of the task. If is the name of the function if the type is a task or an actor task. It is the name of the class if it is a actor scheduling task.

language: str#

The language of the task. E.g., Python, Java, or Cpp.

required_resources: dict#

The required resources to execute the task.

runtime_env_info: str#

The runtime environment information for the task.

parent_task_id: str#

The parent task id. If the parent is a normal task, it will be the task’s id. If the parent runs in a concurrent actor (async actor or threaded actor), it will be the actor’s creation task id.

placement_group_id: str#

The placement group id that’s associated with this task.

worker_id: str#

The worker id that’s associated with this task.

events: List[dict]#

The list of events of the given task. Refer to src/ray/protobuf/common.proto for a detailed explanation of the state breakdowns and typical state transition flow.

profiling_data: List[dict]#

The list of profile events of the given task.

creation_time_ms: Optional[int]#

The time when the task is created. A Unix timestamp in ms.

start_time_ms: Optional[int]#

The time when the task starts to run. A Unix timestamp in ms.

end_time_ms: Optional[int]#

The time when the task is finished or failed. A Unix timestamp in ms.