TaskPool

This class provides an eventlet-based task pool for managing concurrent green threads. It leverages GreenPool to execute tasks asynchronously, supporting dynamic resizing of the pool limit and job termination via greenlet killing. The implementation integrates with system signals to track pool lifecycle events and job execution status.

Attributes

Attribute	Type	Description
Timer	class = Timer	Reference to the Timer class used for scheduling internal events.
signal_safe	boolean = False	Boolean flag indicating if the pool can safely handle POSIX signals, set to False for eventlet.
is_green	boolean = True	Boolean flag indicating that this pool implementation uses green threads.
task_join_will_block	boolean = False	Boolean flag indicating whether joining a task will block the entire process.

---

Constructor

Signature

def TaskPool(
    *args: tuple,
    **kwargs: dict
) - > null

Parameters

Name	Type	Description
*args	tuple	Variable length argument list passed to the base class constructor.
**kwargs	dict	Arbitrary keyword arguments passed to the base class constructor.

---

Methods

---

on_start()

@classmethod
def on_start()

Initializes the Eventlet GreenPool and internal tracking maps, then broadcasts a pool started signal.

---

on_stop()

@classmethod
def on_stop()

Gracefully shuts down the pool by waiting for all active greenlets to complete and sending pre/post shutdown signals.

---

on_apply()

@classmethod
def on_apply(
    target: callable,
    args: list = None,
    kwargs: dict = None,
    callback: callable = None,
    accept_callback: callable = None
)

Schedules a task for execution within the green pool and registers it in the internal process map for tracking.

Parameters

Name	Type	Description
target	callable	The function or task to be executed asynchronously
args	list = None	Positional arguments to pass to the target function
kwargs	dict = None	Keyword arguments to pass to the target function
callback	callable = None	Function to be called with the result upon successful task completion
accept_callback	callable = None	Function to be called when the task is officially accepted by the pool

---

grow()

@classmethod
def grow(
    n: int = 1
)

Increases the maximum number of concurrent greenlets allowed in the pool.

Parameters

Name	Type	Description
n	int = 1	The number of slots to add to the current pool limit

---

shrink()

@classmethod
def shrink(
    n: int = 1
)

Decreases the maximum number of concurrent greenlets allowed in the pool.

Parameters

Name	Type	Description
n	int = 1	The number of slots to remove from the current pool limit

---

terminate_job()

@classmethod
def terminate_job(
    pid: int,
    signal: int = None
)

Forcefully kills a specific greenlet identified by its process ID and waits for it to exit.

Parameters

Name	Type	Description
pid	int	The unique identifier of the greenlet to be terminated
signal	int = None	The signal to send; note that Eventlet greenlets primarily respond to kill() rather than system signals

---