Cache motion-correction results

Hi everyone, I recently noticed that our shared filesystem (which is under a high load when there’s a lot of data processing) is somewhat unreliable: sometimes jobs fail because disk is inaccessible for few minutes. But then, I have to restart a job, even though it has ~30% of images successfully processed.

Is there a way to cache previous results for (at least) motioncorrection jobs? E.g. if I’d press “restart job”, it’d be a way to continue the job, and not start it from the beginning.

This happens automatically. Just mark the failed Patch Motion job as complete - you will have two outputs, one with successful exposures, and one with incomplete exposures. You can then run a second Patch Motion job using the incomplete exposures as input.


thanks! works like a charm :slight_smile: