feat: execute uncorrelated scalar subqueries first instead of rewriting to a join#1912
Open
andygrove wants to merge 10 commits into
Open
feat: execute uncorrelated scalar subqueries first instead of rewriting to a join#1912andygrove wants to merge 10 commits into
andygrove wants to merge 10 commits into
Conversation
Bump the workspace to the published datafusion 54 (and arrow/arrow-flight 58.3, object_store 0.13.2, rustyline 18 in ballista-cli), replacing the branch-54 git-rev pin from the earlier draft. API migration to 54: - as_any removed from ExecutionPlan/DataSource/PhysicalExpr: downcast directly on the trait object (and upcast dyn ShuffleWriter to dyn ExecutionPlan); as_any retained where it still exists (Arrow arrays, UserDefinedLogicalNode, ExtensionOptions). - ExecutionPlan::partition_statistics returns Result<Arc<Statistics>>. - parse_protobuf_partitioning/_hash_partitioning take a PhysicalPlanDecodeContext; serialize_partitioning takes the codec plus the proto converter. - TaskContext::new and FunctionRegistry gain higher-order-function arguments (wired empty in Ballista). - BatchPartitioner::new_hash_partitioner is fallible. Test updates for 54 behavior changes: - approx_percentile_cont_with_weight result; shuffle-writer tests made robust to hash distribution; execution-graph/dot tests use Ballista session config so the small join stays Partitioned; regenerated AQE plan snapshots (projection folded into HashJoinExec).
DataFusion 54's DataSourceExec hands file groups to partition streams from a shared work-queue that is only divided across partitions when all partitions of one plan instance are polled concurrently. Ballista executes one partition per task on its own decoded plan instance, so a task that polls a single partition in isolation drains the whole queue and scans the entire table -- producing N-times-inflated results and N-times the work for every task dispatched after the first wave. Restrict each task's DataSourceExec to the file group for its partition_id before execution (other group slots emptied, partition count preserved) so the lone execute(partition_id) reads only its own group. Skip when partition_id is outside the source's file groups (an operator between the scan and the stage output changed the partition count). Closes apache#1907
…tion DataFusion 54 plans uncorrelated scalar subqueries as a physical ScalarSubqueryExec wrapping a ScalarSubqueryExpr that reads an in-process shared results container. That container cannot cross process or stage boundaries, and datafusion-proto can only deserialize the expr inside its surrounding exec, so when Ballista splits a plan into stages at shuffle boundaries the executor receives a bare ScalarSubqueryExpr and fails to decode it -- hanging queries such as TPC-H q11/q15/q22. Disable datafusion.optimizer.enable_physical_uncorrelated_scalar_subquery in the Ballista restricted configuration so the optimizer rewrites uncorrelated scalar subqueries to joins, which Ballista distributes correctly. Verified via datafusion-cli that the rewrite produces results identical to the physical-execution path. Closes apache#1909
Add unit tests for the two DataFusion 54 distributed-execution fixes: - restrict_scan_to_partition keeps only the task's own file group, leaves the partition count intact, returns None for an out-of-range partition, and ignores non-file-backed plans. - the Ballista restricted configuration disables datafusion.optimizer.enable_physical_uncorrelated_scalar_subquery so uncorrelated scalar subqueries are rewritten to joins.
DataFusion 54 plans an uncorrelated scalar subquery as a physical ScalarSubqueryExec whose ScalarSubqueryExpr cannot be deserialized once Ballista splits the plan into stages. apache#1909 worked around this by decorrelating such subqueries to joins, which is correct but adds join work for what is logically a constant. Instead, in the client query planner, detect uncorrelated scalar subqueries in the optimized logical plan, run each as its own distributed query, reduce the result to a single value (0 rows -> typed null, 1 row -> the value, more than one -> error), and substitute it as a literal before the outer plan is sent to the scheduler. Nested subqueries are inlined recursively. This supersedes the decorrelation workaround, so the Ballista restricted configuration no longer disables datafusion.optimizer.enable_physical_uncorrelated_scalar_subquery. Closes apache#1910
Reconcile this branch's DataFusion 54 work with the canonical DF54 upgrade (apache#1906) that landed upstream: - Keep enable_physical_uncorrelated_scalar_subquery at DataFusion's default (enabled); this branch materializes uncorrelated scalar subqueries in the client planner instead of disabling the option. - Adopt upstream's higher-order function support in the function registry and TaskContext construction. - Adopt upstream's recursive PhysicalPlanDecodeContext codec. - Adopt upstream's more robust restrict_scan_to_partition that handles MemorySourceConfig and warns on unrecognized DataSourceExec sources. - Drop the make_filter_projection_serde_safe workaround (no longer needed on DF54; upstream added regression coverage in serde).
…ze-scalar-subquery # Conflicts: # ballista/core/src/extension.rs
andygrove
marked this pull request as ready for review
July 2, 2026 16:56
Convert the collect/substitute unit tests to assert on the plan rendered with display_indent() instead of counts and emptiness, so each test shows the plan it expects. The skip-correlated test now pairs an eligible subquery with a correlated one and checks the collected plan is the eligible one.
milenkovicm
approved these changes
Jul 11, 2026
milenkovicm
left a comment
milenkovicm
left a comment
Contributor
There was a problem hiding this comment.
thanks @andygrove just few minor comments
| Default::default(), | ||
| ) | ||
| .await?; | ||
|
|
Contributor
There was a problem hiding this comment.
can we explicitly disable SET datafusion.optimizer.enable_physical_uncorrelated_scalar_subquery = true, it would give additional context to the test.
Member
Author
There was a problem hiding this comment.
Done — the test now runs SET datafusion.optimizer.enable_physical_uncorrelated_scalar_subquery = true before the query, so it states that it covers the case where the optimizer is free to plan a physical scalar subquery.
|
|
||
| Ok(()) | ||
| } | ||
|
|
Contributor
There was a problem hiding this comment.
can we also add check with AQE enabled, if it works, if not a jira issue or a test in "unsupported" file?
Member
Author
There was a problem hiding this comment.
Added standalone_aqe / remote_aqe cases (new standalone_context_with_aqe() / remote_context_with_aqe() fixtures set ballista.planner.adaptive.enabled=true). All four cases pass — materialization happens in the client query planner before the plan reaches the scheduler, so it is independent of which distributed planner runs. No issue needed.
…config Set datafusion.optimizer.enable_physical_uncorrelated_scalar_subquery = true in the test so it states the configuration it covers, and add standalone/remote cases with the adaptive planner enabled.
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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.
Which issue does this PR close?
Closes #1910.
Rationale for this change
#1909 made uncorrelated scalar subqueries work in distributed execution by
disabling
datafusion.optimizer.enable_physical_uncorrelated_scalar_subquery,which makes the optimizer rewrite them to joins (DataFusion 54's physical
ScalarSubqueryExeccannot be serialized across Ballista's stage boundaries).That is correct but not efficient: an uncorrelated scalar subquery produces a
single value, but rewriting it to a join turns it into a (cross / left) join
against the subquery's one-row result, adding shuffle/build/probe work for what
is logically a constant.
This PR runs the subquery first and substitutes its value, which is both
efficient and avoids the
ScalarSubqueryExecserialization problem entirely.What changes are included in this PR?
BallistaQueryPlanner::create_physical_plan),before sending the plan to the scheduler:
Expr::ScalarSubquerynodes in the optimizedlogical plan (correlated subqueries are left to DataFusion's existing
decorrelation rules);
single
ScalarValue(0 rows → typed null, 1 row → the value, more than onerow → an error);
materialized recursively, innermost first.
enable_physical_uncorrelated_scalar_subquery = falsedefaultadded in DataFusion 54: uncorrelated scalar subqueries (q11/q15/q22) fail — ScalarSubqueryExpr cannot be deserialized in a split stage plan #1909, since the subqueries are now materialized rather than
decorrelated. This supersedes that workaround.
Are there any user-facing changes?
No API or SQL semantics change. Uncorrelated scalar subqueries (e.g. TPC-H
q11/q15/q22) now execute on a Ballista cluster by computing the subquery first
and inlining its value.
Tested with new unit tests for the detection/substitution helpers (including
that correlated subqueries are skipped) and a standalone + remote integration
test that runs
where id = (select max(id) from test)end to end.cargo test,cargo fmt, andcargo clippy --all-targets -- -D warningsare clean.