Rename bounded_order_preserving_variants config to prefer_exising_sort and update docs
#7723
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
viirya
reviewed
Oct 2, 2023
docs/source/user-guide/configs.md
Outdated
| @@ -87,7 +87,7 @@ Environment variables are read during `SessionConfig` initialisation so they mus | |||
| | datafusion.optimizer.repartition_file_scans | true | When set to `true`, file groups will be repartitioned to achieve maximum parallelism. Currently Parquet and CSV formats are supported. If set to `true`, all files will be repartitioned evenly (i.e., a single large file might be partitioned into smaller chunks) for parallel scanning. If set to `false`, different files will be read in parallel, but repartitioning won't happen within a single file. | | |||
| | datafusion.optimizer.repartition_windows | true | Should DataFusion repartition data using the partitions keys to execute window functions in parallel using the provided `target_partitions` level | | |||
| | datafusion.optimizer.repartition_sorts | true | Should DataFusion execute sorts in a per-partition fashion and merge afterwards instead of coalescing first and sorting globally. With this flag is enabled, plans in the form below `text "SortExec: [a@0 ASC]", " CoalescePartitionsExec", " RepartitionExec: partitioning=RoundRobinBatch(8), input_partitions=1", ` would turn into the plan below which performs better in multithreaded environments `text "SortPreservingMergeExec: [a@0 ASC]", " SortExec: [a@0 ASC]", " RepartitionExec: partitioning=RoundRobinBatch(8), input_partitions=1", ` | | |||
| | datafusion.optimizer.bounded_order_preserving_variants | false | When true, DataFusion will opportunistically remove sorts by replacing `RepartitionExec` with `SortPreservingRepartitionExec`, and `CoalescePartitionsExec` with `SortPreservingMergeExec`, even when the query is bounded. | | |||
| | datafusion.optimizer.prefer_existing_sort | false | When true, DataFusion will opportunistically remove sorts when the data is already sorted, replacing `RepartitionExec` with `SortPreservingRepartitionExec`, and `CoalescePartitionsExec` with `SortPreservingMergeExec`, When false, DataFusion will prefer to maximize the parallelism using `Repartition/Coalesce` and resort the data subsequently with `SortExec` | | |||
There was a problem hiding this comment.
Suggested change
| | datafusion.optimizer.prefer_existing_sort | false | When true, DataFusion will opportunistically remove sorts when the data is already sorted, replacing `RepartitionExec` with `SortPreservingRepartitionExec`, and `CoalescePartitionsExec` with `SortPreservingMergeExec`, When false, DataFusion will prefer to maximize the parallelism using `Repartition/Coalesce` and resort the data subsequently with `SortExec` | | |
| | datafusion.optimizer.prefer_existing_sort | false | When true, DataFusion will opportunistically remove sorts when the data is already sorted, replacing `RepartitionExec` with `SortPreservingRepartitionExec` (i.e., `RepartitionExec` with `preserve_order` as true), and `CoalescePartitionsExec` with `SortPreservingMergeExec`. When false, DataFusion will prefer to maximize the parallelism using `Repartition/Coalesce` and resort the data subsequently with `SortExec` | |
viirya
reviewed
Oct 2, 2023
datafusion/common/src/config.rs
Outdated
| /// When true, DataFusion will opportunistically remove sorts by replacing | ||
| /// `RepartitionExec` with `SortPreservingRepartitionExec`, and | ||
| /// When true, DataFusion will opportunistically remove sorts when the data is already sorted, | ||
| /// replacing `RepartitionExec` with `SortPreservingRepartitionExec`, and |
There was a problem hiding this comment.
Hmm, actually there is no SortPreservingRepartitionExec operator but it is a variant of RepartitionExec with preserve_order as true. It is a little confusion at first if trying to look for SortPreservingRepartitionExec type in IDE.
Maybe:
Suggested change
| /// replacing `RepartitionExec` with `SortPreservingRepartitionExec`, and | |
| /// replacing `RepartitionExec` with `SortPreservingRepartitionExec` (i.e., `RepartitionExec` with `preserve_order` as true), and |
viirya
approved these changes
Oct 2, 2023
Ted-Jiang
pushed a commit
to Ted-Jiang/arrow-datafusion
that referenced
this pull request
Oct 7, 2023
…sort` and update docs (apache#7723) * Improve documentation for bounded_order_preserving_variants config * update docs * fmt * update config * fix typo :facepalm * prettier * Reword for clarity
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.
Which issue does this PR close?
closes #7722
Rationale for this change
While debugging an issue upgrading our code to use DataFuson, @ozankabak pointed me at the following config: #7671 (comment)
This setting (I think) controls if the DataFusion planner should prefer using the existing sort order or trying to maximize paralleilsm using repartition and re-sorting
It turns out to be the right one, but I don't think I would have found it without @ozankabak 's suggestion
I think the core of my challenge is that the current name describes how it modifies DataFusion's algorithms rather than what effect it has on the plans
What changes are included in this PR?
I propose to change the config to
prefer_existing_sortand update the documentationAre these changes tested?
existing tests
Are there any user-facing changes?
yes, a config setting has a different name (and this is a breaking API change)