Support dynamic filters via remote scanner #4077
Open
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
jackkleeman
force-pushed
the
push-ynuqyxwrvmxu
branch
from
249cce6 to
923f158
Compare
Test Results 7 files + 2 7 suites +2 3m 22s ⏱️ + 2m 6s Results for commit 95383dc. ± Comparison against base commit bbf4d42. This pull request removes 34 and adds 47 tests. Note that renamed tests count towards both.♻️ This comment has been updated with latest results. |
jackkleeman
force-pushed
the
push-ynuqyxwrvmxu
branch
from
923f158 to
e9989d8
Compare
igalshilman
approved these changes
Dec 11, 2025
crates/storage-query-datafusion/src/remote_query_scanner_client.rs
Outdated
Show resolved
Hide resolved
crates/storage-query-datafusion/src/remote_query_scanner_client.rs
Outdated
Show resolved
Hide resolved
| if current_predicate_generation != predicate_generation { | ||
| predicate_generation = current_predicate_generation; | ||
| Some(RemoteQueryScannerPredicate { | ||
| serialized_physical_expression: encode_expr(predicate)?, |
Contributor
There was a problem hiding this comment.
- do you know what is the typical size of this, surely small-ish, just making sure.
- do you know how often we are expected to enter that branch? is it configurable?
Contributor
Author
There was a problem hiding this comment.
small yeah - status = "completed" encodes to 33 bytes
i think in the topk case we will probably enter it every batch, as each batch that arrives at the topk exec will update the min or the max and lead to a tighter bound
jackkleeman
force-pushed
the
push-ynuqyxwrvmxu
branch
from
e9989d8 to
30165dc
Compare
jackkleeman
force-pushed
the
push-ynuqyxwrvmxu
branch
from
30165dc to
95383dc
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.
Dynamic filters massively reduce the number of rows returned in topk operations like 'get the 50 most recent invocations'. They are not particularly helpful locally however, as we have to add a row to a record batch to be able to apply a filter, and at this point we have done 99% of the work of the scan. But remotely, they allow us to send less rows over the wire, which is valuable.
In handle_child_pushdown_result we track any dynamic filters and combine them with our initial filter. Then we are able to snapshot the dynamic filters and send them over the wire when they change, so on the remote side a stricter filter can be applied