Skip to content

refactor(query): migrate physical plan from enum to trait-based architecture #18268

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Merged
merged 56 commits into from
Aug 17, 2025
Merged

refactor(query): migrate physical plan from enum to trait-based architecture #18268

merged 56 commits into from
Aug 17, 2025

Conversation

zhang2014
Copy link
Member

@zhang2014 zhang2014 commented Jun 28, 2025 •
edited
Loading

I hereby agree to the terms of the CLA available at: https://docs.databend.com/dev/policies/cla/

Summary

This PR introduces a major architectural refactoring of Databend's physical plan system, migrating from a monolithic enum-based design to a modern trait-based architecture. This change significantly improves
code maintainability, performance, and extensibility while laying the foundation for future query engine enhancements.

Key Changes

Core Architecture Migration

  • Before: Single large PhysicalPlan enum with 50+ variants
  • After: Trait-based system with IPhysicalPlan trait and individual plan types
  • Benefit: Better modularity, easier to extend, and improved compile-time optimization

Trait System Design

  #[typetag::serde]
  pub trait IPhysicalPlan: DynClone + Debug + Send + Sync + 'static {
      fn derive(&self, children: Vec<PhysicalPlan>) -> PhysicalPlan;
      fn build_pipeline2(&self, builder: &mut PipelineBuilder) -> Result<()>;
      fn formatter(&self) -> Result<Box<dyn PhysicalFormat + '_>>;
      // ... other methods
  }

Benefits

  • Easier to add new physical plan types
  • Trait-based design allows for better composition
  • Testability: Each plan can be unit tested independently
  • Code Quality: Better separation of concerns and reduced duplication
  • Consistency: Unified interfaces and formatting across all plans

Tests

  • Unit Test
  • Logic Test
  • Benchmark Test
  • No Test - Explain why

Type of change

  • Bug Fix (non-breaking change which fixes an issue)
  • New Feature (non-breaking change which adds functionality)
  • Breaking Change (fix or feature that could cause existing functionality not to work as expected)
  • Documentation Update
  • Refactoring
  • Performance Improvement
  • Other (please describe):

This change is Reviewable

@github-actions github-actions bot added the pr-refactor this PR changes the code base without new features or bugfix label Jun 28, 2025
forsaken628
forsaken628 reviewed Jul 15, 2025
View reviewed changes
zhang2014 added 15 commits July 15, 2025 13:56
@zhang2014
refactor(query): use trait to refactor physical plan
0dd0362
@zhang2014
refactor(query): use trait to refactor physical plan
81e1b8a
@zhang2014
Merge branch 'main' of https://github.com/datafuselabs/databend into …
373c80d 
…refactor/physical_plan_trait

# Conflicts:
#	src/query/service/src/interpreters/interpreter_copy_into_table.rs
#	src/query/service/src/pipelines/builders/builder_copy_into_location.rs
#	src/query/service/src/pipelines/builders/builder_join.rs
#	src/query/service/src/pipelines/processors/transforms/transform_recursive_cte_source.rs
#	src/query/sql/src/executor/format.rs
#	src/query/sql/src/executor/physical_plans/physical_union_all.rs
@zhang2014
refactor(query): use trait to refactor physical plan
e8a174f
@zhang2014
refactor(query): use trait to refactor physical plan
eb867b1
@zhang2014
refactor(query): use trait to refactor physical plan
66045a3
@zhang2014
refactor(query): use trait to refactor physical plan
9b19b1d
@zhang2014
refactor(query): use trait to refactor physical plan
58186fe
@zhang2014
refactor(query): use trait to refactor physical plan
5e6bedf
@zhang2014
refactor(query): use trait to refactor physical plan
5bd2815
@zhang2014
refactor(query): use trait to refactor physical plan
9f44159
@zhang2014
refactor(query): use trait to refactor physical plan
f0b1997
@zhang2014
Merge branch 'main' of https://github.com/datafuselabs/databend into …
ba7863c 
…refactor/physical_plan_trait

# Conflicts:
#	src/query/service/src/interpreters/interpreter_table_analyze.rs
#	src/query/service/src/physical_plans/physical_limit.rs
#	src/query/service/src/physical_plans/physical_recluster.rs
#	src/query/service/src/physical_plans/physical_table_scan.rs
#	src/query/service/src/pipelines/builders/builder_aggregate.rs
#	src/query/service/src/pipelines/builders/builder_commit.rs
#	src/query/service/src/pipelines/builders/builder_compact.rs
#	src/query/service/src/pipelines/builders/builder_hilbert_partition.rs
#	src/query/service/src/pipelines/builders/builder_row_fetch.rs
#	src/query/service/src/pipelines/builders/builder_sort.rs
#	src/query/service/src/pipelines/builders/builder_union_all.rs
#	src/query/service/src/pipelines/builders/builder_window.rs
#	src/query/service/src/pipelines/builders/mod.rs
#	src/query/service/src/pipelines/pipeline_builder.rs
#	src/query/service/src/pipelines/processors/transforms/transform_recursive_cte_source.rs
#	src/query/service/src/schedulers/fragments/fragmenter.rs
#	src/query/sql/src/executor/format.rs
#	src/query/sql/src/executor/physical_plan.rs
#	src/query/sql/src/executor/physical_plan_visitor.rs
#	src/query/sql/src/executor/physical_plans/mod.rs
#	src/query/sql/src/executor/physical_plans/physical_aggregate_partial.rs
#	src/query/sql/src/executor/physical_plans/physical_sort.rs
#	src/query/sql/src/executor/physical_plans/physical_union_all.rs
#	src/query/sql/src/executor/physical_plans/physical_window_partition.rs
#	src/query/sql/src/planner/optimizer/optimizers/hyper_dp/dynamic_sample/dynamic_sample.rs
@zhang2014
refactor(query): use trait to refactor physical plan
d63e231
@zhang2014
refactor(query): use trait to refactor physical plan
994d159
@zhang2014 zhang2014 marked this pull request as ready for review August 15, 2025 18:55
@zhang2014 zhang2014 added the ci-cloud Build docker image for cloud test label Aug 16, 2025
@zhang2014 zhang2014 requested a review from BohuTANG August 16, 2025 01:17
@github-actions GitHub Actions
Copy link
Contributor

Docker Image for PR

  • tag: pr-18268-3488bd0-1755309078

note: this image tag is only available for internal use.

@zhang2014 zhang2014 changed the title refactor(query): use trait to refactor physical plan refactor(query): migrate physical plan from enum to trait-based architecture Aug 16, 2025
@zhang2014 zhang2014 added the ci-benchmark Benchmark: run all test label Aug 16, 2025
@zhang2014 zhang2014 mentioned this pull request Aug 16, 2025
Merged
11 tasks
@github-actions GitHub Actions
Copy link
Contributor

Docker Image for PR

  • tag: pr-18268-7d0a6b9-1755313503

note: this image tag is only available for internal use.

@github-actions GitHub Actions
Copy link
Contributor

ClickBench Report

@BohuTANG BohuTANG added ci-cloud Build docker image for cloud test and removed ci-cloud Build docker image for cloud test labels Aug 16, 2025
@github-actions GitHub Actions
Copy link
Contributor

Docker Image for PR

  • tag: pr-18268-0cf5486-1755337260

note: this image tag is only available for internal use.

@github-actions GitHub Actions
Copy link
Contributor

ClickBench Report

@zhang2014 zhang2014 force-pushed the refactor/physical_plan_trait branch from a2e3e2d to eea6de3 Compare August 17, 2025 03:13
@zhang2014 zhang2014 merged commit c0f12f9 into databendlabs:main Aug 17, 2025
87 of 88 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@forsaken628 forsaken628 forsaken628 left review comments

@BohuTANG BohuTANG Awaiting requested review from BohuTANG

Assignees
No one assigned
Labels
ci-benchmark Benchmark: run all test ci-cloud Build docker image for cloud test pr-refactor this PR changes the code base without new features or bugfix
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
3 participants
@zhang2014 @forsaken628 @BohuTANG