[SPARK-47746] Implement ordinal-based range encoding in the RocksDBStateEncoder #45905
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
HeartSaVioR
reviewed
Apr 8, 2024
| private val rangeScanKeyFieldsWithIdx: Seq[(StructField, Int)] = { | ||
| keySchema.zipWithIndex.take(numOrderingCols) | ||
| private val rangeScanKeyFieldsWithOrdinal: Seq[(StructField, Int)] = { | ||
| orderingOrdinals.map(ordinal => { |
There was a problem hiding this comment.
| private val remainingKeyFieldsWithIdx: Seq[(StructField, Int)] = { | ||
| keySchema.zipWithIndex.drop(numOrderingCols) | ||
| private val remainingKeyFieldsWithOrdinal: Seq[(StructField, Int)] = { | ||
| 0.to(keySchema.length - 1).diff(orderingOrdinals).map(ordinal => { |
There was a problem hiding this comment.
nit: same, { ordinal =>
| assert(valueRowToData(store.get(keyRow, cfName)) === 1) | ||
| } | ||
|
|
||
| // scalastyle:off |
There was a problem hiding this comment.
Any reason you had to disable style? when you disable style for valid reason, please be specific to disable a single rule, and also make sure to re-enable as long as it doesn't need to be disabled further.
neilramaswamy
force-pushed
the
spark-47746
branch
from
f04801d
to
f6efa6d
Compare
|
Thanks! Merging to master. |
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Outdated suggestions cannot be applied.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.
What changes were proposed in this pull request?
The RocksDBStateEncoder now implements range projection by reading a list of ordering ordinals, and using that to project certain columns, in big-endian, to the front of the
Array[Byte]encoded rows returned by the encoder.Why are the changes needed?
StateV2 implementations (and other state-related operators) project certain columns to the front of
UnsafeRows, and then rely on the RocksDBStateEncoder to range-encode those columns. We can avoid the initial projection by just passing the RocksDBStateEncoder the ordinals to encode at the front. This should avoid any GC or codegen overheads associated with projection.Does this PR introduce any user-facing change?
No.
How was this patch tested?
New UTs. All existing UTs should pass.
Was this patch authored or co-authored using generative AI tooling?
Yes