datafusion 48 #3560
Merged
datafusion 48 #3560
Conversation
This file contains hidden or 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
AdamGS
reviewed
Jun 18, 2025
| log::debug!("DataFusion expression can't be pushed down: {expr:?}"); | ||
| false | ||
| } | ||
| fn can_be_pushed_down(expr: &PhysicalExprRef, schema: &Schema) -> bool { |
There was a problem hiding this comment.
was just thinking last night due we support new things here? We have In now too right? (@joseph-isaacs?)
There was a problem hiding this comment.
But its a pretty slow rn, what type would it run over?
There was a problem hiding this comment.
whatever is available I guess? if its not a perf win right now we can add that later
AdamGS
approved these changes
Jun 18, 2025
|
I think the tests fails because datafusion used to default to |
Benchmarks: TPC-H on NVMETable of Results
|
Benchmarks: TPC-H on S3Table of Results
|
Benchmarks: Clickbench on NVMETable of Results
|
|
IDK what happened to query 6, but we're 20x worse and parquet is 25x worse. There were a bunch of changes to statistics and repartitioning recently and I wonder if that's the issue there, the query is: SELECT MIN("EventDate"), MAX("EventDate") FROM hits; |
|
Wonder if they no longer correctly prune on file level statistics when they added partition group statistics |
|
@robert3005 thats my suspicion, parts of that were already in v47 but v48 has at least one more big change around that |
|
I am also curious, will have a look, query 0 is also 10x @ parquet and 3x @ vortex |
|
I filed an issue upstream apache/datafusion#16444 |
|
I think this fine to merge, we can re-enable the datafusion config (+ #3569) and once apache/datafusion#16447 is released it'll be the default behavior again. |
Signed-off-by: Onur Satici <onur@spiraldb.com>
This is required for apache/datafusion#16447 to take effect (once its merged). By default datafusion will take care of that if you use the SQL interface (and do all the session state setup). Signed-off-by: Adam Gutglick <adam@spiraldb.com>
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.
No description provided.