Skip to content

multiprocessing resume_after can apportion post-resume pipeline tables  #374

@toliwaga

Description

@toliwaga

When resuming a multiprocess run, mp_tasks.apportion will wrongly apportion tables in pipeline created in a later phase of single-process execution. This only happens if in the prior run, execution proceeded through a multiprocess step and coalesced the results back into the pipeline - and then the resume_after step specified repeats the apportion step. I will be comitting a fix for this forthwith.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions