feat: more economic data skipping with datafusion #2772
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
roeap
requested review from
wjones127,
rtyler,
hntd187 and
ion-elgreco
as code owners
roeap
force-pushed
the
feature/data-skipping
branch
from
e1557b5
to
cdbd392
Compare
ion-elgreco
reviewed
Aug 14, 2024
| .ok()?; | ||
| results.push(result.record_batch().clone()); | ||
| } | ||
| let batch = concat_batches(results[0].schema_ref(), &results).ok()?; |
There was a problem hiding this comment.
Isn't it cheaper to concat first and then run it through the Evaluator?
There was a problem hiding this comment.
Probably, at the very least that would give more opportunity for parallelism. in fact we should already concatenate the batches on EagerSnapshot. However the schemas of these batches are not yet normalized, so we cannot concatenate them yet.
For this we need to do either some internal casting/filtering in the log replay, or even better do column selection when reading the checkpoints ...
rtyler
previously approved these changes
Aug 14, 2024
roeap
force-pushed
the
feature/data-skipping
branch
from
928b10c
to
e688ad4
Compare
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Outdated suggestions cannot be applied.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.
Description
This PR adresses some inefficiencies when creating stats for file pruning using the
PruningPredicate. Rather then generating stats from file actions which we generate ad-hoc from teh underlying arrow data, we now read the stats directly from the raw file data - i.e. avoiding expensive roundtrips through file actions.To make this work, we needed to include parsing partition values (adding the
partitionValues_parsedfield to action data) during log replay. As a follow-up we should make more use of the parsed partiton values #2771Related Issue(s)
Documentation