JobLibRunner¶

Defined in scheduler.runners.joblib_runner

A runner that uses joblib for parallel execution. This runner is suitable for local execution with multiple cores. It can handle different job types: - Function: Uses joblib to run Python functions - Script: Executes scripts in separate processes - Container: Runs containers using Docker or Singularity

Inherits from: BaseRunner

Class Definition¶

class JobLibRunner(self, n_jobs: <class 'int'> = -1, backend: <class 'str'> = loky, config: Dict[str, Any] = None):
    """
    Initialize a new JobLibRunner.
    **Args:**
    * **n_jobs**: Number of jobs to run in parallel (-1 for all cores)
    * **backend**: Backend to use for joblib (loky, threading, multiprocessing)
    * **config**: Additional configuration options:
    * **container_engine**: 'docker' or 'singularity' (default: 'docker')
    * **tmp_dir**: Directory for temporary files (default: system temp dir)
    """

Methods¶

Method	Description
`cancel_job`	Cancel a job.
`check_job_status`	Check the status of a job and update its state.
`run_job`	Run a job using the appropriate execution method.
`shutdown`	Shutdown the executor.

Method Details¶

cancel_job¶

def cancel_job(self, job: Any) -> None

Cancel a job. Args: * job: The job to cancel

check_job_status¶

def check_job_status(self, job: Any) -> None

Check the status of a job and update its state. Args: * job: The job to check

run_job¶

def run_job(self, job: Any) -> None

Run a job using the appropriate execution method. Args: * job: The job to run

shutdown¶

def shutdown(self) -> Any

Shutdown the executor.