This repository has been archived by the owner on Sep 18, 2023. It is now read-only.
[NSE-400] Native Arrow Row to columnar support #637
Merged
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
|
Thanks for opening a pull request! Could you open an issue for this pull request on Github Issues? https://github.com/oap-project/native-sql-engine/issues Then could you also rename commit message and pull request title in the following format? See also: |
zhouyuan
changed the title
…RowToColumnarExec
|
@haojinIntel can you do a rebase? the format check is fixed in master |
zhouyuan
reviewed
Jan 5, 2022
| val res = new Iterator[ColumnarBatch] { | ||
| private val converters = new RowToColumnConverter(localSchema) | ||
| private var last_cb: ColumnarBatch = null | ||
| private var elapse: Long = 0 |
zhouyuan
reviewed
Jan 5, 2022
| // Allocate large buffer to store the numRows rows | ||
| val bufferSize = 134217728 // 128M can estimator the buffer size based on the data type | ||
| val allocator = SparkMemoryUtils.contextAllocator() | ||
| val arrowBuf: ArrowBuf = allocator.buffer(bufferSize) |
There was a problem hiding this comment.
reuse buf allocated in #95
zhouyuan
reviewed
Jan 5, 2022
| } | ||
| val timeZoneId = SparkSchemaUtils.getLocalTimezoneID() | ||
| val arrowSchema = ArrowUtils.toArrowSchema(schema, timeZoneId) | ||
| val schemaBytes: Array[Byte] = ConverterUtils.getSchemaBytesBuf(arrowSchema) |
There was a problem hiding this comment.
move these varaiables to init
zhouyuan
reviewed
Jan 5, 2022
| out_data.null_count = null_count; | ||
| *array = MakeArray(std::make_shared<arrow::ArrayData>(std::move(out_data))); | ||
| return arrow::Status::OK(); | ||
| } else if (type->id() == arrow::Int8Type::type_id) { |
zhouyuan
changed the title
zhouyuan
changed the title
zhouyuan
added a commit
to zhouyuan/native-sql-engine
that referenced
this pull request
Jan 8, 2022
* Support ArrowRowToColumnar Optimization * Replace expired code * Add the code to convert recordbatch to columnarBatch * Add unit test on java size * Update the unit tests * Fix the bug when reading decimal value from unsafeRow * Use ArrowRowToColumnarExec instead of RowToArrowColumnarExec * Use clang-format to standardize the CPP code format * enable arrowRowToColumnarExec * Add the metrics for ArrowRowToColumnarExec * Add the metrics for ArrowRowToColumnarExec and unsupport Codegen * Add parameter 'spark.oap.sql.columnar.rowtocolumnar' to control ArrowRowToColumnarExec * Remove useless code * Release arrowbuf after return recordbatch * Fix the processTime metric for ArrowRowToColumnarExec * Refine the code of ArrowRowToColumnar operator * Add more metrics to detect the elapse time of each action * Small fix about allocating buffer for unsafeRow * Remove useless code * Remove useless metrics for ArrowRowToColumnarExec * Fall back to use java RowToColumnarExec when the row is not unsafeRow Type * Fix the bug for decimal format * fix format Co-authored-by: Yuan Zhou <yuan.zhou@intel.com>
zhouyuan
added a commit
that referenced
this pull request
Jan 9, 2022
* Use Hadoop 3.2 as default hadoop.version (#652) * [NSE-661] Add left/right trim in WSCG * [NSE-675] Add instr expression support (#676) * Initial commit * Add the support in wscg * [NSE-674] Add translate expression support (#672) * Initial commit * Add StringTranslate for subquery checking * Code refactor * Change arrow branch for unit test [will revert at last] * Revert "Change arrow branch for unit test [will revert at last]" This reverts commit bf74356. * Port the function to wscg * Change arrow branch for unit test [will revert at last] * Format native code * Fix a bug * Revert "Change arrow branch for unit test [will revert at last]" This reverts commit 3a53fa2. * [NSE-681] Add floor & ceil expression support (#682) * Initial commit * Add ceil expression support * Change arrow branch for unit test [will revert at last] * Revert "Change arrow branch for unit test [will revert at last]" This reverts commit 5fb2f4b. * [NSE-647] Leverage buffered write in shuffle (#648) Closes #647 * [NSE-400] Native Arrow Row to columnar support (#637) * Support ArrowRowToColumnar Optimization * Replace expired code * Add the code to convert recordbatch to columnarBatch * Add unit test on java size * Update the unit tests * Fix the bug when reading decimal value from unsafeRow * Use ArrowRowToColumnarExec instead of RowToArrowColumnarExec * Use clang-format to standardize the CPP code format * enable arrowRowToColumnarExec * Add the metrics for ArrowRowToColumnarExec * Add the metrics for ArrowRowToColumnarExec and unsupport Codegen * Add parameter 'spark.oap.sql.columnar.rowtocolumnar' to control ArrowRowToColumnarExec * Remove useless code * Release arrowbuf after return recordbatch * Fix the processTime metric for ArrowRowToColumnarExec * Refine the code of ArrowRowToColumnar operator * Add more metrics to detect the elapse time of each action * Small fix about allocating buffer for unsafeRow * Remove useless code * Remove useless metrics for ArrowRowToColumnarExec * Fall back to use java RowToColumnarExec when the row is not unsafeRow Type * Fix the bug for decimal format * fix format Co-authored-by: Yuan Zhou <yuan.zhou@intel.com> * fix leakage in rowtocolumn (#683) Signed-off-by: Yuan Zhou <yuan.zhou@intel.com> Co-authored-by: Wei-Ting Chen <weiting.chen@intel.com> Co-authored-by: PHILO-HE <feilong.he@intel.com> Co-authored-by: Hongze Zhang <hongze.zhang@intel.com> Co-authored-by: haojinIntel <hao.jin@intel.com>
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
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?
Native Arrow Row to columnar support
How was this patch tested?
pass jenkins