Ensure DAG-level references are filled on unmap #33083
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Previously, a serialized mapped operator's unmapped SerializedOperator misses some DAG-level references because they were established separately in BaseSerialization and were overlooked when the unmap() function was implemented. This extracts those post-population ref-fixing code into a function, and call it as needed in unmap(), so an unmapped SerializedOperator is consistent with a non-mapped SerializedOperator that comes straightly out of a database. This was not an issue prior to 2.6 since the scheduler mostly did not access DAG-level references on a serialized operator (mapped or not). The introduction of the fail_fast flag requires accessing the DAG much later in a task's lifetime in the scheduler, and thus needs the references to be properly set.
uranusjr
force-pushed
the
serialized-mapped-operator-unmap-dag
branch
from
c44021e
to
812f81b
Compare
jedcunningham
approved these changes
Aug 3, 2023
ephraimbuddy
pushed a commit
that referenced
this pull request
Aug 4, 2023
Co-authored-by: Jed Cunningham <66968678+jedcunningham@users.noreply.github.com> (cherry picked from commit bcfadcf)
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Previously, a serialized mapped operator's unmapped SerializedOperator misses some DAG-level references because they were established separately in BaseSerialization and were overlooked when the unmap() function was implemented. This extracts those post-population ref-fixing code into a function, and call it as needed in unmap(), so an unmapped SerializedOperator is consistent with a non-mapped SerializedOperator that comes straightly out of a database.
This was not an issue prior to 2.6 since the scheduler mostly did not access DAG-level references on a serialized operator (mapped or not). The introduction of the fail_fast flag requires accessing the DAG much later in a task's lifetime in the scheduler, and thus needs the references to be properly set.