orca.tasks.imaging_tasks

Imaging Celery tasks for snapshot and batch imaging.

Provides Celery tasks for imaging pipelines including: - Single measurement set imaging with calibration - Batch imaging with shared NVMe workspace - PNG preview generation from FITS images

Attributes

logger
NVME_ROOT

Functions

imaging_pipeline_task(→ str)	Runs copy→applycal→WSClean→PNG→remove extra files→export→purge NVMe
imaging_shared_pipeline_task(→ str)	Shared-workspace version:
run_image_pipeline(→ Dict[str, str])

Module Contents

orca.tasks.imaging_tasks.logger

orca.tasks.imaging_tasks.NVME_ROOT = '/fast/pipeline'

orca.tasks.imaging_tasks.imaging_pipeline_task(ms_path: str, delay_table: str, bandpass_table: str, final_dir: str, extra_wsclean: List[str] = None) → str

Runs copy→applycal→WSClean→PNG→remove extra files→export→purge NVMe and returns the path to the saved PNG.

orca.tasks.imaging_tasks.imaging_shared_pipeline_task(self, ms_path: str, delay_table: str, bandpass_table: str, final_dir: str, workdir_root: str, keep_full_products: bool = False, extra_wsclean: List[str] | None = None, bad_corrs: List[int] | None = None, peel_sources: str | None = None) → str

Shared-workspace version:

• All MSs in a batch share workdir_root.

• Only the first MS (or any with keep_full_products=True) keeps the full WSClean output; others retain just -dirty FITS + PNG.

• Automatic retries wipe the MS sub-dir first to avoid half-baked data.

orca.tasks.imaging_tasks.run_image_pipeline(self, ms_path: str, workdir: str, dest: str) → Dict[str, str]

1. Images one Measurement Set.

2. Moves the temporary workdir to Lustre on the same node.

Parameters:

• ms_path (str) – Absolute path to the averaged *.ms file.

• workdir (str) – Fast scratch / NVMe directory to use for intermediate files.

• dest (str) – Final directory on Lustre where the whole workdir is moved.

Returns:

{“fits”: <final FITS file>, “dest”: <final workdir location>}

Return type:

dict