perf: Improve benchmarks for native row-to-columnar used by JVM shuffle #3290
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
Add jvm_shuffle.rs benchmark that covers the full range of data types processed by `process_sorted_row_partition()` in JVM shuffle: - Primitive columns (100 Int64 columns) - Struct (flat with 5/10/20 fields) - Nested struct (2 levels deep) - Deeply nested struct (3 levels deep) - List<Int64> - Map<Int64, Int64> This replaces the old row_columnar.rs which only tested primitive columns. These benchmarks help measure the performance of the row-to-columnar conversion used by CometColumnarShuffle when writing shuffle data. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## main #3290 +/- ##
============================================
+ Coverage 56.12% 60.13% +4.00%
- Complexity 976 1468 +492
============================================
Files 119 175 +56
Lines 11743 16085 +4342
Branches 2251 2665 +414
============================================
+ Hits 6591 9672 +3081
- Misses 4012 5066 +1054
- Partials 1140 1347 +207 ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
Use div_ceil() instead of manual ceiling division and replace needless range loop with iterator pattern. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
andygrove
changed the title
comphead
reviewed
Jan 27, 2026
| fn get_row_size(num_struct_fields: usize) -> usize { | ||
| // Top-level row has 1 column (the struct) | ||
| let top_level_bitset_width = SparkUnsafeRow::get_row_bitset_width(1); | ||
| // Struct pointer (offset + size) is 8 bytes |
Contributor
There was a problem hiding this comment.
Suggested change
| // Struct pointer (offset + size) is 8 bytes | |
| // Struct pointer (offset + size) is 8 bytes on 64bit architectures |
comphead
reviewed
Jan 27, 2026
| let top_level_bitset_width = SparkUnsafeRow::get_row_bitset_width(1); | ||
|
|
||
| // Nested struct starts after top-level row header + pointer | ||
| let nested_offset = top_level_bitset_width + 8; |
Contributor
There was a problem hiding this comment.
just a thought, its too many eights, prob it would be easy to name them? where is the pointer size or int64 size, etc?
comphead
reviewed
Jan 27, 2026
| // Fill nested struct with some data | ||
| for i in 0..num_struct_fields { | ||
| let value_offset = nested_offset + nested_bitset_width + i * 8; | ||
| let value = (i as i64) * 100; |
Contributor
There was a problem hiding this comment.
what is 100 here? is aligning?
comphead
reviewed
Jan 27, 2026
| false, | ||
| 0, | ||
| None, | ||
| &CompressionCodec::Zstd(1), |
Contributor
There was a problem hiding this comment.
should we also check other codecs? 🤔
I might be wrong but Spark uses LZ4 for IO_COMPRESSION_CODEC which is used for shuffle?
private[spark] val IO_COMPRESSION_CODEC =
ConfigBuilder("spark.io.compression.codec")
.doc("The codec used to compress internal data such as RDD partitions, event log, " +
"broadcast variables and shuffle outputs. By default, Spark provides four codecs: " +
"lz4, lzf, snappy, and zstd. You can also use fully qualified class names to specify " +
"the codec")
.version("0.8.0")
.stringConf
.createWithDefaultString("lz4")
comphead
reviewed
Jan 27, 2026
| } | ||
|
|
||
| /// Create a schema with nested structs: Struct<Struct<int64 fields>> | ||
| fn make_nested_struct_schema(num_fields: usize) -> DataType { |
Contributor
There was a problem hiding this comment.
I have some feeling make_nested_struct_schema and make_deeply_nested_struct_schema can be generalized?
comphead
reviewed
Jan 27, 2026
| fn get_nested_row_size(num_inner_fields: usize) -> usize { | ||
| // Top-level row has 1 column (the outer struct) | ||
| let top_level_bitset_width = SparkUnsafeRow::get_row_bitset_width(1); | ||
| let struct_pointer_size = 8; |
Contributor
There was a problem hiding this comment.
this value should prob be const on crate level?
comphead
reviewed
Jan 27, 2026
Contributor
comphead
left a comment
comphead
left a comment
There was a problem hiding this comment.
Thanks @andygrove WDYT it would be nice to get a followup PR to see nested type of list/map/struct combinations?
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 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.
Improve
row_columnar.rsbenchmark to cover the full range of data types processed byprocess_sorted_row_partition()in JVM shuffle:These benchmarks help measure the performance of the row-to-columnar conversion used by CometColumnarShuffle when writing shuffle data.