## What changes were proposed in this pull request?
`colStat.min` AND `colStat.max` are empty for string type. Thus, `evaluateInSet` should not return zero when either `colStat.min` or `colStat.max`.
## How was this patch tested?
Added a test case.
Author: gatorsmile <gatorsmile@gmail.com>
Closes#21147 from gatorsmile/cached.
## What changes were proposed in this pull request?
The PR adds the SQL function `array_join`. The behavior of the function is based on Presto's one.
The function accepts an `array` of `string` which is to be joined, a `string` which is the delimiter to use between the items of the first argument and optionally a `string` which is used to replace `null` values.
## How was this patch tested?
added UTs
Author: Marco Gaido <marcogaido91@gmail.com>
Closes#21011 from mgaido91/SPARK-23916.
## What changes were proposed in this pull request?
HIVE-15511 introduced the `roundOff` flag in order to disable the rounding to 8 digits which is performed in `months_between`. Since this can be a computational intensive operation, skipping it may improve performances when the rounding is not needed.
## How was this patch tested?
modified existing UT
Author: Marco Gaido <marcogaido91@gmail.com>
Closes#21008 from mgaido91/SPARK-23902.
## What changes were proposed in this pull request?
a followup of https://github.com/apache/spark/pull/21100
## How was this patch tested?
N/A
Author: Wenchen Fan <wenchen@databricks.com>
Closes#21154 from cloud-fan/test.
## What changes were proposed in this pull request?
Union of map and other compatible column result in unresolved operator 'Union; exception
Reproduction
`spark-sql>select map(1,2), 'str' union all select map(1,2,3,null), 1`
Output:
```
Error in query: unresolved operator 'Union;;
'Union
:- Project [map(1, 2) AS map(1, 2)#106, str AS str#107]
: +- OneRowRelation$
+- Project [map(1, cast(2 as int), 3, cast(null as int)) AS map(1, CAST(2 AS INT), 3, CAST(NULL AS INT))#109, 1 AS 1#108]
+- OneRowRelation$
```
So, we should cast part of columns to be compatible when appropriate.
## How was this patch tested?
Added a test (query union of map and other columns) to SQLQueryTestSuite's union.sql.
Author: liutang123 <liutang123@yeah.net>
Closes#21100 from liutang123/SPARK-24012.
## What changes were proposed in this pull request?
This pr is a follow-up of #20980 and fixes code to reuse `InternalRow` for converting input keys/values in `ExternalMapToCatalyst` eval.
## How was this patch tested?
Existing tests.
Author: Takeshi Yamamuro <yamamuro@apache.org>
Closes#21137 from maropu/SPARK-23589-FOLLOWUP.
## What changes were proposed in this pull request?
In SPARK-23375 we introduced the ability of removing `Sort` operation during query optimization if the data is already sorted. In this follow-up we remove also a `Sort` which is followed by another `Sort`: in this case the first sort is not needed and can be safely removed.
The PR starts from henryr's comment: https://github.com/apache/spark/pull/20560#discussion_r180601594. So credit should be given to him.
## How was this patch tested?
added UT
Author: Marco Gaido <marcogaido91@gmail.com>
Closes#21072 from mgaido91/SPARK-23973.
## What changes were proposed in this pull request?
This pr supported interpreted mode for `ExternalMapToCatalyst`.
## How was this patch tested?
Added tests in `ObjectExpressionsSuite`.
Author: Takeshi Yamamuro <yamamuro@apache.org>
Closes#20980 from maropu/SPARK-23589.
## What changes were proposed in this pull request?
The existing query constraints framework has 2 steps:
1. propagate constraints bottom up.
2. use constraints to infer additional filters for better data pruning.
For step 2, it mostly helps with Join, because we can connect the constraints from children to the join condition and infer powerful filters to prune the data of the join sides. e.g., the left side has constraints `a = 1`, the join condition is `left.a = right.a`, then we can infer `right.a = 1` to the right side and prune the right side a lot.
However, the current logic of inferring filters from constraints for Join is pretty weak. It infers the filters from Join's constraints. Some joins like left semi/anti exclude output from right side and the right side constraints will be lost here.
This PR propose to check the left and right constraints individually, expand the constraints with join condition and add filters to children of join directly, instead of adding to the join condition.
This reverts https://github.com/apache/spark/pull/20670 , covers https://github.com/apache/spark/pull/20717 and https://github.com/apache/spark/pull/20816
This is inspired by the original PRs and the tests are all from these PRs. Thanks to the authors mgaido91 maryannxue KaiXinXiaoLei !
## How was this patch tested?
new tests
Author: Wenchen Fan <wenchen@databricks.com>
Closes#21083 from cloud-fan/join.
## What changes were proposed in this pull request?
A followup of https://github.com/apache/spark/pull/20988
`PhysicalOperation` can collect Project and Filters over a certain plan and substitute the alias with the original attributes in the bottom plan. We can use it in `OptimizeMetadataOnlyQuery` rule to handle the Project and Filter over partitioned relation.
## How was this patch tested?
existing test
Author: Wenchen Fan <wenchen@databricks.com>
Closes#21111 from cloud-fan/refactor.
>What changes were proposed in this pull request?
During evaluation of IN conditions, if the source data frame, is represented by a plan, that uses hive table with columns, which were previously analysed, and the plan has conditions for these fields, that cannot be satisfied (which leads us to an empty data frame), FilterEstimation.evaluateInSet method produces NumberFormatException and ClassCastException.
In order to fix this bug, method FilterEstimation.evaluateInSet at first checks, if distinct count is not zero, and also checks if colStat.min and colStat.max are defined, and only in this case proceeds with the calculation. If at least one of the conditions is not satisfied, zero is returned.
>How was this patch tested?
In order to test the PR two tests were implemented: one in FilterEstimationSuite, that tests the plan with the statistics that violates the conditions mentioned above, and another one in StatisticsCollectionSuite, that test the whole process of analysis/optimisation of the query, that leads to the problems, mentioned in the first section.
Author: Mykhailo Shtelma <mykhailo.shtelma@bearingpoint.com>
Author: smikesh <mshtelma@gmail.com>
Closes#21052 from mshtelma/filter_estimation_evaluateInSet_Bugs.
## What changes were proposed in this pull request?
When the OffsetWindowFunction's frame is `UnaryMinus(Literal(1))` but the specified window frame has been simplified to `Literal(-1)` by some optimizer rules e.g., `ConstantFolding`. Thus, they do not match and cause the following error:
```
org.apache.spark.sql.AnalysisException: Window Frame specifiedwindowframe(RowFrame, -1, -1) must match the required frame specifiedwindowframe(RowFrame, -1, -1);
at org.apache.spark.sql.catalyst.analysis.CheckAnalysis$class.failAnalysis(CheckAnalysis.scala:41)
at org.apache.spark.sql.catalyst.analysis.Analyzer.failAnalysis(Analyzer.scala:91)
at
```
## How was this patch tested?
Added a test
Author: gatorsmile <gatorsmile@gmail.com>
Closes#21115 from gatorsmile/fixLag.
## What changes were proposed in this pull request?
This pr supported interpreted mode for `ValidateExternalType`.
## How was this patch tested?
Added tests in `ObjectExpressionsSuite`.
Author: Takeshi Yamamuro <yamamuro@apache.org>
Closes#20757 from maropu/SPARK-23595.
## What changes were proposed in this pull request?
This pr is a follow-up pr of #20979 and fixes code to resolve a map builder method per execution instead of per row in `CatalystToExternalMap`.
## How was this patch tested?
Existing tests.
Author: Takeshi Yamamuro <yamamuro@apache.org>
Closes#21112 from maropu/SPARK-23588-FOLLOWUP.
## What changes were proposed in this pull request?
Currently we find the wider common type by comparing the two types from left to right, this can be a problem when you have two data types which don't have a common type but each can be promoted to StringType.
For instance, if you have a table with the schema:
[c1: date, c2: string, c3: int]
The following succeeds:
SELECT coalesce(c1, c2, c3) FROM table
While the following produces an exception:
SELECT coalesce(c1, c3, c2) FROM table
This is only a issue when the seq of dataTypes contains `StringType` and all the types can do string promotion.
close#19033
## How was this patch tested?
Add test in `TypeCoercionSuite`
Author: Xingbo Jiang <xingbo.jiang@databricks.com>
Closes#21074 from jiangxb1987/typeCoercion.
## What changes were proposed in this pull request?
This pr supported interpreted mode for `CatalystToExternalMap`.
## How was this patch tested?
Added tests in `ObjectExpressionsSuite`.
Author: Takeshi Yamamuro <yamamuro@apache.org>
Closes#20979 from maropu/SPARK-23588.
## What changes were proposed in this pull request?
This pr supported interpreted mode for `NewInstance`.
## How was this patch tested?
Added tests in `ObjectExpressionsSuite`.
Author: Takeshi Yamamuro <yamamuro@apache.org>
Closes#20778 from maropu/SPARK-23584.
## What changes were proposed in this pull request?
The PR adds the SQL function `element_at`. The behavior of the function is based on Presto's one.
This function returns element of array at given index in value if column is array, or returns value for the given key in value if column is map.
## How was this patch tested?
Added UTs
Author: Kazuaki Ishizaki <ishizaki@jp.ibm.com>
Closes#21053 from kiszk/SPARK-23924.
## What changes were proposed in this pull request?
The PR adds the SQL function `array_position`. The behavior of the function is based on Presto's one.
The function returns the position of the first occurrence of the element in array x (or 0 if not found) using 1-based index as BigInt.
## How was this patch tested?
Added UTs
Author: Kazuaki Ishizaki <ishizaki@jp.ibm.com>
Closes#21037 from kiszk/SPARK-23919.
## What changes were proposed in this pull request?
Use specified accessor in `ArrayData.foreach` and `toArray`.
## How was this patch tested?
Existing tests.
Author: Liang-Chi Hsieh <viirya@gmail.com>
Closes#21099 from viirya/SPARK-23875-followup.
## What changes were proposed in this pull request?
`EqualNullSafe` for `FloatType` and `DoubleType` might generate a wrong result by codegen.
```scala
scala> val df = Seq((Some(-1.0d), None), (None, Some(-1.0d))).toDF()
df: org.apache.spark.sql.DataFrame = [_1: double, _2: double]
scala> df.show()
+----+----+
| _1| _2|
+----+----+
|-1.0|null|
|null|-1.0|
+----+----+
scala> df.filter("_1 <=> _2").show()
+----+----+
| _1| _2|
+----+----+
|-1.0|null|
|null|-1.0|
+----+----+
```
The result should be empty but the result remains two rows.
## How was this patch tested?
Added a test.
Author: Takuya UESHIN <ueshin@databricks.com>
Closes#21094 from ueshin/issues/SPARK-24007/equalnullsafe.
## What changes were proposed in this pull request?
Added `TransitPredicateInOuterJoin` optimization rule that transits constraints from the preserved side of an outer join to the null-supplying side. The constraints of the join operator will remain unchanged.
## How was this patch tested?
Added 3 tests in `InferFiltersFromConstraintsSuite`.
Author: maryannxue <maryann.xue@gmail.com>
Closes#20816 from maryannxue/spark-21479.
## What changes were proposed in this pull request?
We are using `CodegenContext.freshName` to get a unique name for any new variable we are adding. Unfortunately, this method currently fails to create a unique name when we request more than one instance of variables with starting name `name1` and an instance with starting name `name11`.
The PR changes the way a new name is generated by `CodegenContext.freshName` so that we generate unique names in this scenario too.
## How was this patch tested?
added UT
Author: Marco Gaido <marcogaido91@gmail.com>
Closes#21080 from mgaido91/SPARK-23986.
## What changes were proposed in this pull request?
In current code, it will scanning all partition paths when spark.sql.hive.verifyPartitionPath=true.
e.g. table like below:
```
CREATE TABLE `test`(
`id` int,
`age` int,
`name` string)
PARTITIONED BY (
`A` string,
`B` string)
load data local inpath '/tmp/data0' into table test partition(A='00', B='00')
load data local inpath '/tmp/data1' into table test partition(A='01', B='01')
load data local inpath '/tmp/data2' into table test partition(A='10', B='10')
load data local inpath '/tmp/data3' into table test partition(A='11', B='11')
```
If I query with SQL – "select * from test where A='00' and B='01' ", current code will scan all partition paths including '/data/A=00/B=00', '/data/A=00/B=00', '/data/A=01/B=01', '/data/A=10/B=10', '/data/A=11/B=11'. It costs much time and memory cost.
This pr proposes to avoid iterating all partition paths. Add a config `spark.files.ignoreMissingFiles` and ignore the `file not found` when `getPartitions/compute`(for hive table scan). This is much like the logic brought by
`spark.sql.files.ignoreMissingFiles`(which is for datasource scan).
## How was this patch tested?
UT
Author: jinxing <jinxing6042@126.com>
Closes#19868 from jinxing64/SPARK-22676.
## What changes were proposed in this pull request?
There was no check on nullability for arguments of `Tuple`s. This could lead to have weird behavior when a null value had to be deserialized into a non-nullable Scala object: in those cases, the `null` got silently transformed in a valid value (like `-1` for `Int`), corresponding to the default value we are using in the SQL codebase. This situation was very likely to happen when deserializing to a Tuple of primitive Scala types (like Double, Int, ...).
The PR adds the `AssertNotNull` to arguments of tuples which have been asked to be converted to non-nullable types.
## How was this patch tested?
added UT
Author: Marco Gaido <marcogaido91@gmail.com>
Closes#20976 from mgaido91/SPARK-23835.
## What changes were proposed in this pull request?
We don't have a good way to sequentially access `UnsafeArrayData` with a common interface such as `Seq`. An example is `MapObject` where we need to access several sequence collection types together. But `UnsafeArrayData` doesn't implement `ArrayData.array`. Calling `toArray` will copy the entire array. We can provide an `IndexedSeq` wrapper for `ArrayData`, so we can avoid copying the entire array.
## How was this patch tested?
Added test.
Author: Liang-Chi Hsieh <viirya@gmail.com>
Closes#20984 from viirya/SPARK-23875.
## What changes were proposed in this pull request?
The PR adds the SQL function `array_min`. It takes an array as argument and returns the minimum value in it.
## How was this patch tested?
added UTs
Author: Marco Gaido <marcogaido91@gmail.com>
Closes#21025 from mgaido91/SPARK-23918.
## What changes were proposed in this pull request?
Currently, interpreted execution of `LambdaVariable` just uses `InternalRow.get` to access element. We should use specified accessors if possible.
## How was this patch tested?
Added test.
Author: Liang-Chi Hsieh <viirya@gmail.com>
Closes#20981 from viirya/SPARK-23873.
## What changes were proposed in this pull request?
The PR adds the SQL function `array_max`. It takes an array as argument and returns the maximum value in it.
## How was this patch tested?
added UTs
Author: Marco Gaido <marcogaido91@gmail.com>
Closes#21024 from mgaido91/SPARK-23917.
## What changes were proposed in this pull request?
Just found `MultiAlias` is a `CodegenFallback`. It should not be as looks like `MultiAlias` won't be evaluated.
## How was this patch tested?
Existing tests.
Author: Liang-Chi Hsieh <viirya@gmail.com>
Closes#21065 from viirya/multialias-without-codegenfallback.
## What changes were proposed in this pull request?
Checkpoint files (offset log files, state store files) in Structured Streaming must be written atomically such that no partial files are generated (would break fault-tolerance guarantees). Currently, there are 3 locations which try to do this individually, and in some cases, incorrectly.
1. HDFSOffsetMetadataLog - This uses a FileManager interface to use any implementation of `FileSystem` or `FileContext` APIs. It preferably loads `FileContext` implementation as FileContext of HDFS has atomic renames.
1. HDFSBackedStateStore (aka in-memory state store)
- Writing a version.delta file - This uses FileSystem APIs only to perform a rename. This is incorrect as rename is not atomic in HDFS FileSystem implementation.
- Writing a snapshot file - Same as above.
#### Current problems:
1. State Store behavior is incorrect - HDFS FileSystem implementation does not have atomic rename.
1. Inflexible - Some file systems provide mechanisms other than write-to-temp-file-and-rename for writing atomically and more efficiently. For example, with S3 you can write directly to the final file and it will be made visible only when the entire file is written and closed correctly. Any failure can be made to terminate the writing without making any partial files visible in S3. The current code does not abstract out this mechanism enough that it can be customized.
#### Solution:
1. Introduce a common interface that all 3 cases above can use to write checkpoint files atomically.
2. This interface must provide the necessary interfaces that allow customization of the write-and-rename mechanism.
This PR does that by introducing the interface `CheckpointFileManager` and modifying `HDFSMetadataLog` and `HDFSBackedStateStore` to use the interface. Similar to earlier `FileManager`, there are implementations based on `FileSystem` and `FileContext` APIs, and the latter implementation is preferred to make it work correctly with HDFS.
The key method this interface has is `createAtomic(path, overwrite)` which returns a `CancellableFSDataOutputStream` that has the method `cancel()`. All users of this method need to either call `close()` to successfully write the file, or `cancel()` in case of an error.
## How was this patch tested?
New tests in `CheckpointFileManagerSuite` and slightly modified existing tests.
Author: Tathagata Das <tathagata.das1565@gmail.com>
Closes#21048 from tdas/SPARK-23966.
## What changes were proposed in this pull request?
Add UDF weekday
## How was this patch tested?
A new test
Author: yucai <yyu1@ebay.com>
Closes#21009 from yucai/SPARK-23905.
## What changes were proposed in this pull request?
Current SS continuous doesn't support processing on temp table or `df.as("xxx")`, SS will throw an exception as LogicalPlan not supported, details described in [here](https://issues.apache.org/jira/browse/SPARK-23748).
So here propose to add this support.
## How was this patch tested?
new UT.
Author: jerryshao <sshao@hortonworks.com>
Closes#21017 from jerryshao/SPARK-23748.
## What changes were proposed in this pull request?
This PR tries to use `MemoryBlock` in `UTF8StringBuffer`. In general, there are two advantages to use `MemoryBlock`.
1. Has clean API calls rather than using a Java array or `PlatformMemory`
2. Improve runtime performance of memory access instead of using `Object`.
## How was this patch tested?
Added `UTF8StringBufferSuite`
Author: Kazuaki Ishizaki <ishizaki@jp.ibm.com>
Closes#20871 from kiszk/SPARK-23762.
## What changes were proposed in this pull request?
This PR slightly refactors the newly added `ExprValue` API by quite a bit. The following changes are introduced:
1. `ExprValue` now uses the actual class instead of the class name as its type. This should give some more flexibility with generating code in the future.
2. Renamed `StatementValue` to `SimpleExprValue`. The statement concept is broader then an expression (untyped and it cannot be on the right hand side of an assignment), and this was not really what we were using it for. I have added a top level `JavaCode` trait that can be used in the future to reinstate (no pun intended) a statement a-like code fragment.
3. Added factory methods to the `JavaCode` companion object to make it slightly less verbose to create `JavaCode`/`ExprValue` objects. This is also what makes the diff quite large.
4. Added one more factory method to `ExprCode` to make it easier to create code-less expressions.
## How was this patch tested?
Existing tests.
Author: Herman van Hovell <hvanhovell@databricks.com>
Closes#21026 from hvanhovell/SPARK-23951.
## What changes were proposed in this pull request?
In the PR #20886, I mistakenly check the table location only when `ignoreIfExists` is false, which was following the original deprecated PR.
That was wrong. When `ignoreIfExists` is true and the target table doesn't exist, we should also check the table location. In other word, **`ignoreIfExists` has nothing to do with table location validation**.
This is a follow-up PR to fix the mistake.
## How was this patch tested?
Add one unit test.
Author: Gengliang Wang <gengliang.wang@databricks.com>
Closes#21001 from gengliangwang/SPARK-19724-followup.
## What changes were proposed in this pull request?
This PR moves writing of `UnsafeRow`, `UnsafeArrayData` & `UnsafeMapData` out of the `GenerateUnsafeProjection`/`InterpretedUnsafeProjection` classes into the `UnsafeWriter` interface. This cleans up the code a little bit, and it should also result in less byte code for the code generated path.
## How was this patch tested?
Existing tests
Author: Herman van Hovell <hvanhovell@databricks.com>
Closes#20986 from hvanhovell/SPARK-23864.
## What changes were proposed in this pull request?
Code generation for the `Add` and `Subtract` expressions was not done using the `BinaryArithmetic.doCodeGen` method because these expressions also support `CalendarInterval`. This leads to a bit of duplication.
This PR gets rid of that duplication by adding `calendarIntervalMethod` to `BinaryArithmetic` and doing the code generation for `CalendarInterval` in `BinaryArithmetic` instead.
## How was this patch tested?
Existing tests.
Author: Herman van Hovell <hvanhovell@databricks.com>
Closes#21005 from hvanhovell/SPARK-23898.
## What changes were proposed in this pull request?
Add `hashUTF8String()` to the hasher classes to allow Spark SQL codegen to generate cleaner code for hashing `UTF8String`s. No change in behavior otherwise.
Although with the introduction of SPARK-10399, the code size for hashing `UTF8String` is already smaller, it's still good to extract a separate function in the hasher classes so that the generated code can stay clean.
## How was this patch tested?
Existing tests.
Author: Kris Mok <kris.mok@databricks.com>
Closes#21016 from rednaxelafx/hashutf8.
## What changes were proposed in this pull request?
The codegen output of `Expression`, aka `ExprCode`, now encapsulates only strings of output value (`value`) and nullability (`isNull`). It makes difficulty for us to know what the output really is. I think it is better if we can add wrappers for the value and nullability that let us to easily know that.
## How was this patch tested?
Existing tests.
Author: Liang-Chi Hsieh <viirya@gmail.com>
Closes#20043 from viirya/SPARK-22856.
## What changes were proposed in this pull request?
This PR avoids possible overflow at an operation `long = (long)(int * int)`. The multiplication of large positive integer values may set one to MSB. This leads to a negative value in long while we expected a positive value (e.g. `0111_0000_0000_0000 * 0000_0000_0000_0010`).
This PR performs long cast before the multiplication to avoid this situation.
## How was this patch tested?
Existing UTs
Author: Kazuaki Ishizaki <ishizaki@jp.ibm.com>
Closes#21002 from kiszk/SPARK-23893.
## What changes were proposed in this pull request?
This PR is to finish https://github.com/apache/spark/pull/17272
This JIRA is a follow up work after SPARK-19583
As we discussed in that PR
The following DDL for a managed table with an existed default location should throw an exception:
CREATE TABLE ... (PARTITIONED BY ...) AS SELECT ...
CREATE TABLE ... (PARTITIONED BY ...)
Currently there are some situations which are not consist with above logic:
CREATE TABLE ... (PARTITIONED BY ...) succeed with an existed default location
situation: for both hive/datasource(with HiveExternalCatalog/InMemoryCatalog)
CREATE TABLE ... (PARTITIONED BY ...) AS SELECT ...
situation: hive table succeed with an existed default location
This PR is going to make above two situations consist with the logic that it should throw an exception
with an existed default location.
## How was this patch tested?
unit test added
Author: Gengliang Wang <gengliang.wang@databricks.com>
Closes#20886 from gengliangwang/pr-17272.
Fixes https://issues.apache.org/jira/browse/SPARK-23823
Keep origin for all the methods using transformExpression
## What changes were proposed in this pull request?
Keep origin in transformExpression
## How was this patch tested?
Manually tested that this fixes https://issues.apache.org/jira/browse/SPARK-23823 and columns have correct origins after Analyzer.analyze
Author: JiahuiJiang <jjiang@palantir.com>
Author: Jiahui Jiang <jjiang@palantir.com>
Closes#20961 from JiahuiJiang/jj/keep-origin.
## What changes were proposed in this pull request?
This PR allows us to use one of several types of `MemoryBlock`, such as byte array, int array, long array, or `java.nio.DirectByteBuffer`. To use `java.nio.DirectByteBuffer` allows to have off heap memory which is automatically deallocated by JVM. `MemoryBlock` class has primitive accessors like `Platform.getInt()`, `Platform.putint()`, or `Platform.copyMemory()`.
This PR uses `MemoryBlock` for `OffHeapColumnVector`, `UTF8String`, and other places. This PR can improve performance of operations involving memory accesses (e.g. `UTF8String.trim`) by 1.8x.
For now, this PR does not use `MemoryBlock` for `BufferHolder` based on cloud-fan's [suggestion](https://github.com/apache/spark/pull/11494#issuecomment-309694290).
Since this PR is a successor of #11494, close#11494. Many codes were ported from #11494. Many efforts were put here. **I think this PR should credit to yzotov.**
This PR can achieve **1.1-1.4x performance improvements** for operations in `UTF8String` or `Murmur3_x86_32`. Other operations are almost comparable performances.
Without this PR
```
OpenJDK 64-Bit Server VM 1.8.0_121-8u121-b13-0ubuntu1.16.04.2-b13 on Linux 4.4.0-22-generic
Intel(R) Xeon(R) CPU E5-2667 v3 3.20GHz
OpenJDK 64-Bit Server VM 1.8.0_121-8u121-b13-0ubuntu1.16.04.2-b13 on Linux 4.4.0-22-generic
Intel(R) Xeon(R) CPU E5-2667 v3 3.20GHz
Hash byte arrays with length 268435487: Best/Avg Time(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------
Murmur3_x86_32 526 / 536 0.0 131399881.5 1.0X
UTF8String benchmark: Best/Avg Time(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------
hashCode 525 / 552 1022.6 1.0 1.0X
substring 414 / 423 1298.0 0.8 1.3X
```
With this PR
```
OpenJDK 64-Bit Server VM 1.8.0_121-8u121-b13-0ubuntu1.16.04.2-b13 on Linux 4.4.0-22-generic
Intel(R) Xeon(R) CPU E5-2667 v3 3.20GHz
Hash byte arrays with length 268435487: Best/Avg Time(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------
Murmur3_x86_32 474 / 488 0.0 118552232.0 1.0X
UTF8String benchmark: Best/Avg Time(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------
hashCode 476 / 480 1127.3 0.9 1.0X
substring 287 / 291 1869.9 0.5 1.7X
```
Benchmark program
```
test("benchmark Murmur3_x86_32") {
val length = 8192 * 32768 + 31
val seed = 42L
val iters = 1 << 2
val random = new Random(seed)
val arrays = Array.fill[MemoryBlock](numArrays) {
val bytes = new Array[Byte](length)
random.nextBytes(bytes)
new ByteArrayMemoryBlock(bytes, Platform.BYTE_ARRAY_OFFSET, length)
}
val benchmark = new Benchmark("Hash byte arrays with length " + length,
iters * numArrays, minNumIters = 20)
benchmark.addCase("HiveHasher") { _: Int =>
var sum = 0L
for (_ <- 0L until iters) {
sum += HiveHasher.hashUnsafeBytesBlock(
arrays(i), Platform.BYTE_ARRAY_OFFSET, length)
}
}
benchmark.run()
}
test("benchmark UTF8String") {
val N = 512 * 1024 * 1024
val iters = 2
val benchmark = new Benchmark("UTF8String benchmark", N, minNumIters = 20)
val str0 = new java.io.StringWriter() { { for (i <- 0 until N) { write(" ") } } }.toString
val s0 = UTF8String.fromString(str0)
benchmark.addCase("hashCode") { _: Int =>
var h: Int = 0
for (_ <- 0L until iters) { h += s0.hashCode }
}
benchmark.addCase("substring") { _: Int =>
var s: UTF8String = null
for (_ <- 0L until iters) { s = s0.substring(N / 2 - 5, N / 2 + 5) }
}
benchmark.run()
}
```
I run [this benchmark program](https://gist.github.com/kiszk/94f75b506c93a663bbbc372ffe8f05de) using [the commit](ee5a79861c). I got the following results:
```
OpenJDK 64-Bit Server VM 1.8.0_151-8u151-b12-0ubuntu0.16.04.2-b12 on Linux 4.4.0-66-generic
Intel(R) Xeon(R) CPU E5-2667 v3 3.20GHz
Memory access benchmarks: Best/Avg Time(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------
ByteArrayMemoryBlock get/putInt() 220 / 221 609.3 1.6 1.0X
Platform get/putInt(byte[]) 220 / 236 610.9 1.6 1.0X
Platform get/putInt(Object) 492 / 494 272.8 3.7 0.4X
OnHeapMemoryBlock get/putLong() 322 / 323 416.5 2.4 0.7X
long[] 221 / 221 608.0 1.6 1.0X
Platform get/putLong(long[]) 321 / 321 418.7 2.4 0.7X
Platform get/putLong(Object) 561 / 563 239.2 4.2 0.4X
```
I also run [this benchmark program](https://gist.github.com/kiszk/5fdb4e03733a5d110421177e289d1fb5) for comparing performance of `Platform.copyMemory()`.
```
OpenJDK 64-Bit Server VM 1.8.0_151-8u151-b12-0ubuntu0.16.04.2-b12 on Linux 4.4.0-66-generic
Intel(R) Xeon(R) CPU E5-2667 v3 3.20GHz
Platform copyMemory: Best/Avg Time(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------
Object to Object 1961 / 1967 8.6 116.9 1.0X
System.arraycopy Object to Object 1917 / 1921 8.8 114.3 1.0X
byte array to byte array 1961 / 1968 8.6 116.9 1.0X
System.arraycopy byte array to byte array 1909 / 1937 8.8 113.8 1.0X
int array to int array 1921 / 1990 8.7 114.5 1.0X
double array to double array 1918 / 1923 8.7 114.3 1.0X
Object to byte array 1961 / 1967 8.6 116.9 1.0X
Object to short array 1965 / 1972 8.5 117.1 1.0X
Object to int array 1910 / 1915 8.8 113.9 1.0X
Object to float array 1971 / 1978 8.5 117.5 1.0X
Object to double array 1919 / 1944 8.7 114.4 1.0X
byte array to Object 1959 / 1967 8.6 116.8 1.0X
int array to Object 1961 / 1970 8.6 116.9 1.0X
double array to Object 1917 / 1924 8.8 114.3 1.0X
```
These results show three facts:
1. According to the second/third or sixth/seventh results in the first experiment, if we use `Platform.get/putInt(Object)`, we achieve more than 2x worse performance than `Platform.get/putInt(byte[])` with concrete type (i.e. `byte[]`).
2. According to the second/third or fourth/fifth/sixth results in the first experiment, the fastest way to access an array element on Java heap is `array[]`. **Cons of `array[]` is that it is not possible to support unaligned-8byte access.**
3. According to the first/second/third or fourth/sixth/seventh results in the first experiment, `getInt()/putInt() or getLong()/putLong()` in subclasses of `MemoryBlock` can achieve comparable performance to `Platform.get/putInt()` or `Platform.get/putLong()` with concrete type (second or sixth result). There is no overhead regarding virtual call.
4. According to results in the second experiment, for `Platform.copy()`, to pass `Object` can achieve the same performance as to pass any type of primitive array as source or destination.
5. According to second/fourth results in the second experiment, `Platform.copy()` can achieve the same performance as `System.arrayCopy`. **It would be good to use `Platform.copy()` since `Platform.copy()` can take any types for src and dst.**
We are incrementally replace `Platform.get/putXXX` with `MemoryBlock.get/putXXX`. This is because we have two advantages.
1) Achieve better performance due to having a concrete type for an array.
2) Use simple OO design instead of passing `Object`
It is easy to use `MemoryBlock` in `InternalRow`, `BufferHolder`, `TaskMemoryManager`, and others that are already abstracted. It is not easy to use `MemoryBlock` in utility classes related to hashing or others.
Other candidates are
- UnsafeRow, UnsafeArrayData, UnsafeMapData, SpecificUnsafeRowJoiner
- UTF8StringBuffer
- BufferHolder
- TaskMemoryManager
- OnHeapColumnVector
- BytesToBytesMap
- CachedBatch
- classes for hash
- others.
## How was this patch tested?
Added `UnsafeMemoryAllocator`
Author: Kazuaki Ishizaki <ishizaki@jp.ibm.com>
Closes#19222 from kiszk/SPARK-10399.
## What changes were proposed in this pull request?
Add interpreted execution for `InitializeJavaBean` expression.
## How was this patch tested?
Added unit test.
Author: Liang-Chi Hsieh <viirya@gmail.com>
Closes#20985 from viirya/SPARK-23593-2.
## What changes were proposed in this pull request?
This pr added interpreted execution for `StaticInvoke`.
## How was this patch tested?
Added tests in `ObjectExpressionsSuite`.
Author: Kazuaki Ishizaki <ishizaki@jp.ibm.com>
Closes#20753 from kiszk/SPARK-23582.
## What changes were proposed in this pull request?
Add interpreted execution for `InitializeJavaBean` expression.
## How was this patch tested?
Added unit test.
Author: Liang-Chi Hsieh <viirya@gmail.com>
Closes#20756 from viirya/SPARK-23593.
## What changes were proposed in this pull request?
This pr added interpreted execution for `Invoke`.
## How was this patch tested?
Added tests in `ObjectExpressionsSuite`.
Author: Kazuaki Ishizaki <ishizaki@jp.ibm.com>
Closes#20797 from kiszk/SPARK-28583.
## What changes were proposed in this pull request?
Add cast to nulls introduced by PropagateEmptyRelation so in cases they're part of coalesce they will not break its type checking rules
## How was this patch tested?
Added unit test
Author: Robert Kruszewski <robertk@palantir.com>
Closes#20914 from robert3005/rk/propagate-empty-fix.
## What changes were proposed in this pull request?
Add interpreted execution for `MapObjects` expression.
## How was this patch tested?
Added unit test.
Author: Liang-Chi Hsieh <viirya@gmail.com>
Closes#20771 from viirya/SPARK-23587.
## What changes were proposed in this pull request?
This PR implemented the following cleanups related to `UnsafeWriter` class:
- Remove code duplication between `UnsafeRowWriter` and `UnsafeArrayWriter`
- Make `BufferHolder` class internal by delegating its accessor methods to `UnsafeWriter`
- Replace `UnsafeRow.setTotalSize(...)` with `UnsafeRowWriter.setTotalSize()`
## How was this patch tested?
Tested by existing UTs
Author: Kazuaki Ishizaki <ishizaki@jp.ibm.com>
Closes#20850 from kiszk/SPARK-23713.
## What changes were proposed in this pull request?
This PR is to improve the test coverage of the original PR https://github.com/apache/spark/pull/20687
## How was this patch tested?
N/A
Author: gatorsmile <gatorsmile@gmail.com>
Closes#20911 from gatorsmile/addTests.
## What changes were proposed in this pull request?
This PR supports for pushing down filters for DateType in parquet
## How was this patch tested?
Added UT and tested in local.
Author: yucai <yyu1@ebay.com>
Closes#20851 from yucai/SPARK-23727.
## What changes were proposed in this pull request?
This PR proposes to add lineSep option for a configurable line separator in text datasource.
It supports this option by using `LineRecordReader`'s functionality with passing it to the constructor.
The approach is similar with https://github.com/apache/spark/pull/20727; however, one main difference is, it uses text datasource's `lineSep` option to parse line by line in JSON's schema inference.
## How was this patch tested?
Manually tested and unit tests were added.
Author: hyukjinkwon <gurwls223@apache.org>
Author: hyukjinkwon <gurwls223@gmail.com>
Closes#20877 from HyukjinKwon/linesep-json.
## What changes were proposed in this pull request?
The UUID() expression is stateful and should implement the `Stateful` trait instead of the `Nondeterministic` trait.
## How was this patch tested?
Added test.
Author: Liang-Chi Hsieh <viirya@gmail.com>
Closes#20912 from viirya/SPARK-23794.
## What changes were proposed in this pull request?
This PR fixes an incorrect comparison in SQL between timestamp and date. This is because both of them are casted to `string` and then are compared lexicographically. This implementation shows `false` regarding this query `spark.sql("select cast('2017-03-01 00:00:00' as timestamp) between cast('2017-02-28' as date) and cast('2017-03-01' as date)").show`.
This PR shows `true` for this query by casting `date("2017-03-01")` to `timestamp("2017-03-01 00:00:00")`.
(Please fill in changes proposed in this fix)
## How was this patch tested?
Added new UTs to `TypeCoercionSuite`.
Author: Kazuaki Ishizaki <ishizaki@jp.ibm.com>
Closes#20774 from kiszk/SPARK-23549.
## What changes were proposed in this pull request?
As stated in Jira, there are problems with current `Uuid` expression which uses `java.util.UUID.randomUUID` for UUID generation.
This patch uses the newly added `RandomUUIDGenerator` for UUID generation. So we can make `Uuid` deterministic between retries.
## How was this patch tested?
Added unit tests.
Author: Liang-Chi Hsieh <viirya@gmail.com>
Closes#20861 from viirya/SPARK-23599-2.
## What changes were proposed in this pull request?
Fixed `CodegenContext.withSubExprEliminationExprs()` so that it saves/restores CSE state correctly.
## How was this patch tested?
Added new unit test to verify that the old CSE state is indeed saved and restored around the `withSubExprEliminationExprs()` call. Manually verified that this test fails without this patch.
Author: Kris Mok <kris.mok@databricks.com>
Closes#20870 from rednaxelafx/codegen-subexpr-fix.
## What changes were proposed in this pull request?
To drop `exprId`s for `Alias` in user-facing info., this pr added an entry for `Alias` in `NonSQLExpression.sql`
## How was this patch tested?
Added tests in `UDFSuite`.
Author: Takeshi Yamamuro <yamamuro@apache.org>
Closes#20827 from maropu/SPARK-23666.
## What changes were proposed in this pull request?
Complex type simplification optimizer rules were not applied to the
entire plan, just the expressions reachable from the root node. This
patch fixes the rules to transform the entire plan.
## How was this patch tested?
New unit test + ran sql / core tests.
Author: Henry Robinson <henry@apache.org>
Author: Henry Robinson <henry@cloudera.com>
Closes#20687 from henryr/spark-25000.
## What changes were proposed in this pull request?
This patch adds a UUID generator from Pseudo-Random Numbers. We can use it later to have deterministic `UUID()` expression.
## How was this patch tested?
Added unit tests.
Author: Liang-Chi Hsieh <viirya@gmail.com>
Closes#20817 from viirya/SPARK-23599.
## What changes were proposed in this pull request?
We currently can only create unsafe rows using code generation. This is a problem for situations in which code generation fails. There is no fallback, and as a result we cannot execute the query.
This PR adds an interpreted version of `UnsafeProjection`. The implementation is modeled after `InterpretedMutableProjection`. It stores the expression results in a `GenericInternalRow`, and it then uses a conversion function to convert the `GenericInternalRow` into an `UnsafeRow`.
This PR does not implement the actual code generated to interpreted fallback logic. This will be done in a follow-up.
## How was this patch tested?
I am piggybacking on exiting `UnsafeProjection` tests, and I have added an interpreted version for each of these.
Author: Herman van Hovell <hvanhovell@databricks.com>
Closes#20750 from hvanhovell/SPARK-23581.
## What changes were proposed in this pull request?
This PR enables assertions in `XXH64Suite.testKnownByteArrayInputs()` on big endian platform, too. The current implementation performs them only on little endian platform. This PR increase test coverage of big endian platform.
## How was this patch tested?
Updated `XXH64Suite`
Tested on big endian platform using JIT compiler or interpreter `-Xint`.
Author: Kazuaki Ishizaki <ishizaki@jp.ibm.com>
Closes#20804 from kiszk/SPARK-23656.
## What changes were proposed in this pull request?
The error message ```s"""Field "$name" does not exist."""``` is thrown when looking up an unknown field in StructType. In the error message, we should also contain the information about which columns/fields exist in this struct.
## How was this patch tested?
Added new unit tests.
Note: I created a new `StructTypeSuite.scala` as I couldn't find an existing suite that's suitable to place these tests. I may be missing something so feel free to propose new locations.
Please review http://spark.apache.org/contributing.html before opening a pull request.
Author: Xiayun Sun <xiayunsun@gmail.com>
Closes#20649 from xysun/SPARK-23462.
## What changes were proposed in this pull request?
The from_json() function accepts an additional parameter, where the user might specify the schema. The issue is that the specified schema might not be compatible with data. In particular, the JSON data might be missing data for fields declared as non-nullable in the schema. The from_json() function does not verify the data against such errors. When data with missing fields is sent to the parquet encoder, there is no verification either. The end results is a corrupt parquet file.
To avoid corruptions, make sure that all fields in the user-specified schema are set to be nullable.
Since this changes the behavior of a public function, we need to include it in release notes.
The behavior can be reverted by setting `spark.sql.fromJsonForceNullableSchema=false`
## How was this patch tested?
Added two new tests.
Author: Michał Świtakowski <michal.switakowski@databricks.com>
Closes#20694 from mswit-databricks/SPARK-23173.
## What changes were proposed in this pull request?
`PrintToStderr` was doing what is it supposed to only when code generation is enabled.
The PR adds the same behavior in interpreted mode too.
## How was this patch tested?
added UT
Author: Marco Gaido <marcogaido91@gmail.com>
Closes#20773 from mgaido91/SPARK-23602.
## What changes were proposed in this pull request?
There was a bug in `calculateParamLength` which caused it to return always 1 + the number of expressions. This could lead to Exceptions especially with expressions of type long.
## How was this patch tested?
added UT + fixed previous UT
Author: Marco Gaido <marcogaido91@gmail.com>
Closes#20772 from mgaido91/SPARK-23628.
## What changes were proposed in this pull request?
The PR adds interpreted execution to DecodeUsingSerializer.
## How was this patch tested?
added UT
Please review http://spark.apache.org/contributing.html before opening a pull request.
Author: Marco Gaido <marcogaido91@gmail.com>
Closes#20760 from mgaido91/SPARK-23592.
## What changes were proposed in this pull request?
This PR adds a configuration to control the fallback of Arrow optimization for `toPandas` and `createDataFrame` with Pandas DataFrame.
## How was this patch tested?
Manually tested and unit tests added.
You can test this by:
**`createDataFrame`**
```python
spark.conf.set("spark.sql.execution.arrow.enabled", False)
pdf = spark.createDataFrame([[{'a': 1}]]).toPandas()
spark.conf.set("spark.sql.execution.arrow.enabled", True)
spark.conf.set("spark.sql.execution.arrow.fallback.enabled", True)
spark.createDataFrame(pdf, "a: map<string, int>")
```
```python
spark.conf.set("spark.sql.execution.arrow.enabled", False)
pdf = spark.createDataFrame([[{'a': 1}]]).toPandas()
spark.conf.set("spark.sql.execution.arrow.enabled", True)
spark.conf.set("spark.sql.execution.arrow.fallback.enabled", False)
spark.createDataFrame(pdf, "a: map<string, int>")
```
**`toPandas`**
```python
spark.conf.set("spark.sql.execution.arrow.enabled", True)
spark.conf.set("spark.sql.execution.arrow.fallback.enabled", True)
spark.createDataFrame([[{'a': 1}]]).toPandas()
```
```python
spark.conf.set("spark.sql.execution.arrow.enabled", True)
spark.conf.set("spark.sql.execution.arrow.fallback.enabled", False)
spark.createDataFrame([[{'a': 1}]]).toPandas()
```
Author: hyukjinkwon <gurwls223@gmail.com>
Closes#20678 from HyukjinKwon/SPARK-23380-conf.
## What changes were proposed in this pull request?
The PR adds interpreted execution to EncodeUsingSerializer.
## How was this patch tested?
added UT
Author: Marco Gaido <marcogaido91@gmail.com>
Closes#20751 from mgaido91/SPARK-23591.
## What changes were proposed in this pull request?
This pr added a helper function in `ExpressionEvalHelper` to check exceptions in all the path of expression evaluation.
## How was this patch tested?
Modified the existing tests.
Author: Takeshi Yamamuro <yamamuro@apache.org>
Closes#20748 from maropu/SPARK-23611.
## What changes were proposed in this pull request?
The PR adds interpreted execution to CreateExternalRow
## How was this patch tested?
added UT
Author: Marco Gaido <marcogaido91@gmail.com>
Closes#20749 from mgaido91/SPARK-23590.
## What changes were proposed in this pull request?
This pr added interpreted execution for `GetExternalRowField`.
## How was this patch tested?
Added tests in `ObjectExpressionsSuite`.
Author: Takeshi Yamamuro <yamamuro@apache.org>
Closes#20746 from maropu/SPARK-23594.
## What changes were proposed in this pull request?
The PR adds interpreted execution to WrapOption.
## How was this patch tested?
added UT
Author: Marco Gaido <marcogaido91@gmail.com>
Closes#20741 from mgaido91/SPARK-23586_2.
## What changes were proposed in this pull request?
The PR adds interpreted execution to UnwrapOption.
## How was this patch tested?
added UT
Author: Marco Gaido <marcogaido91@gmail.com>
Closes#20736 from mgaido91/SPARK-23586.
## What changes were proposed in this pull request?
Provide more details in trigonometric function documentations. Referenced `java.lang.Math` for further details in the descriptions.
## How was this patch tested?
Ran full build, checked generated documentation manually
Author: Mihaly Toth <misutoth@gmail.com>
Closes#20618 from misutoth/trigonometric-doc.
## What changes were proposed in this pull request?
A current `CodegenContext` class has immutable value or method without mutable state, too.
This refactoring moves them to `CodeGenerator` object class which can be accessed from anywhere without an instantiated `CodegenContext` in the program.
## How was this patch tested?
Existing tests
Author: Kazuaki Ishizaki <ishizaki@jp.ibm.com>
Closes#20700 from kiszk/SPARK-23546.
## What changes were proposed in this pull request?
It looks like this was incorrectly copied from `XPathFloat` in the class above.
## How was this patch tested?
(Please explain how this patch was tested. E.g. unit tests, integration tests, manual tests)
(If this patch involves UI changes, please attach a screenshot; otherwise, remove this)
Please review http://spark.apache.org/contributing.html before opening a pull request.
Author: Eric Liang <ekhliang@gmail.com>
Closes#20730 from ericl/fix-typo-xpath.
## What changes were proposed in this pull request?
https://github.com/apache/spark/pull/18944 added one patch, which allowed a spark session to be created when the hive metastore server is down. However, it did not allow running any commands with the spark session. This brings troubles to the user who only wants to read / write data frames without metastore setup.
## How was this patch tested?
Added some unit tests to read and write data frames based on the original HiveMetastoreLazyInitializationSuite.
Please review http://spark.apache.org/contributing.html before opening a pull request.
Author: Feng Liu <fengliu@databricks.com>
Closes#20681 from liufengdb/completely-lazy.
## What changes were proposed in this pull request?
(Please fill in changes proposed in this fix)
I run a sql: `select ls.cs_order_number from ls left semi join catalog_sales cs on ls.cs_order_number = cs.cs_order_number`, The `ls` table is a small table ,and the number is one. The `catalog_sales` table is a big table, and the number is 10 billion. The task will be hang up. And i find the many null values of `cs_order_number` in the `catalog_sales` table. I think the null value should be removed in the logical plan.
>== Optimized Logical Plan ==
>Join LeftSemi, (cs_order_number#1 = cs_order_number#22)
>:- Project cs_order_number#1
> : +- Filter isnotnull(cs_order_number#1)
> : +- MetastoreRelation 100t, ls
>+- Project cs_order_number#22
> +- MetastoreRelation 100t, catalog_sales
Now, use this patch, the plan will be:
>== Optimized Logical Plan ==
>Join LeftSemi, (cs_order_number#1 = cs_order_number#22)
>:- Project cs_order_number#1
> : +- Filter isnotnull(cs_order_number#1)
> : +- MetastoreRelation 100t, ls
>+- Project cs_order_number#22
> : **+- Filter isnotnull(cs_order_number#22)**
> :+- MetastoreRelation 100t, catalog_sales
## How was this patch tested?
(Please explain how this patch was tested. E.g. unit tests, integration tests, manual tests)
(If this patch involves UI changes, please attach a screenshot; otherwise, remove this)
Please review http://spark.apache.org/contributing.html before opening a pull request.
Author: KaiXinXiaoLei <584620569@qq.com>
Author: hanghang <584620569@qq.com>
Closes#20670 from KaiXinXiaoLei/Spark-23405.
## What changes were proposed in this pull request?
Clarify JSON and CSV reader behavior in document.
JSON doesn't support partial results for corrupted records.
CSV only supports partial results for the records with more or less tokens.
## How was this patch tested?
Pass existing tests.
Author: Liang-Chi Hsieh <viirya@gmail.com>
Closes#20666 from viirya/SPARK-23448-2.
## What changes were proposed in this pull request?
```Scala
val tablePath = new File(s"${path.getCanonicalPath}/cOl3=c/cOl1=a/cOl5=e")
Seq(("a", "b", "c", "d", "e")).toDF("cOl1", "cOl2", "cOl3", "cOl4", "cOl5")
.write.json(tablePath.getCanonicalPath)
val df = spark.read.json(path.getCanonicalPath).select("CoL1", "CoL5", "CoL3").distinct()
df.show()
```
It generates a wrong result.
```
[c,e,a]
```
We have a bug in the rule `OptimizeMetadataOnlyQuery `. We should respect the attribute order in the original leaf node. This PR is to fix it.
## How was this patch tested?
Added a test case
Author: gatorsmile <gatorsmile@gmail.com>
Closes#20684 from gatorsmile/optimizeMetadataOnly.
## What changes were proposed in this pull request?
Refactor ColumnStat to be more flexible.
* Split `ColumnStat` and `CatalogColumnStat` just like `CatalogStatistics` is split from `Statistics`. This detaches how the statistics are stored from how they are processed in the query plan. `CatalogColumnStat` keeps `min` and `max` as `String`, making it not depend on dataType information.
* For `CatalogColumnStat`, parse column names from property names in the metastore (`KEY_VERSION` property), not from metastore schema. This means that `CatalogColumnStat`s can be created for columns even if the schema itself is not stored in the metastore.
* Make all fields optional. `min`, `max` and `histogram` for columns were optional already. Having them all optional is more consistent, and gives flexibility to e.g. drop some of the fields through transformations if they are difficult / impossible to calculate.
The added flexibility will make it possible to have alternative implementations for stats, and separates stats collection from stats and estimation processing in plans.
## How was this patch tested?
Refactored existing tests to work with refactored `ColumnStat` and `CatalogColumnStat`.
New tests added in `StatisticsSuite` checking that backwards / forwards compatibility is not broken.
Author: Juliusz Sompolski <julek@databricks.com>
Closes#20624 from juliuszsompolski/SPARK-23445.
## What changes were proposed in this pull request?
This PR always adds `codegenStageId` in comment of the generated class. This is a replication of #20419 for post-Spark 2.3.
Closes#20419
```
/* 001 */ public Object generate(Object[] references) {
/* 002 */ return new GeneratedIteratorForCodegenStage1(references);
/* 003 */ }
/* 004 */
/* 005 */ // codegenStageId=1
/* 006 */ final class GeneratedIteratorForCodegenStage1 extends org.apache.spark.sql.execution.BufferedRowIterator {
/* 007 */ private Object[] references;
...
```
## How was this patch tested?
Existing tests
Author: Kazuaki Ishizaki <ishizaki@jp.ibm.com>
Closes#20612 from kiszk/SPARK-23424.
## What changes were proposed in this pull request?
Apache Spark 2.3 introduced `native` ORC supports with vectorization and many fixes. However, it's shipped as a not-default option. This PR enables `native` ORC implementation and predicate-pushdown by default for Apache Spark 2.4. We will improve and stabilize ORC data source before Apache Spark 2.4. And, eventually, Apache Spark will drop old Hive-based ORC code.
## How was this patch tested?
Pass the Jenkins with existing tests.
Author: Dongjoon Hyun <dongjoon@apache.org>
Closes#20634 from dongjoon-hyun/SPARK-23456.
## What changes were proposed in this pull request?
Cleaned up the codegen templates for `Literal`s, to make sure that the `ExprCode` returned from `Literal.doGenCode()` has:
1. an empty `code` field;
2. an `isNull` field of either literal `true` or `false`;
3. a `value` field that is just a simple literal/constant.
Before this PR, there are a couple of paths that would return a non-trivial `code` and all of them are actually unnecessary. The `NaN` and `Infinity` constants for `double` and `float` can be accessed through constants directly available so there's no need to add a reference for them.
Also took the opportunity to add a new util method for ease of creating `ExprCode` for inline-able non-null values.
## How was this patch tested?
Existing tests.
Author: Kris Mok <kris.mok@databricks.com>
Closes#20626 from rednaxelafx/codegen-literal.
## What changes were proposed in this pull request?
Migrating KafkaSource (with data source v1) to KafkaMicroBatchReader (with data source v2).
Performance comparison:
In a unit test with in-process Kafka broker, I tested the read throughput of V1 and V2 using 20M records in a single partition. They were comparable.
## How was this patch tested?
Existing tests, few modified to be better tests than the existing ones.
Author: Tathagata Das <tathagata.das1565@gmail.com>
Closes#20554 from tdas/SPARK-23362.
## What changes were proposed in this pull request?
To prevent any regressions, this PR changes ORC implementation to `hive` by default like Spark 2.2.X.
Users can enable `native` ORC. Also, ORC PPD is also restored to `false` like Spark 2.2.X.
## How was this patch tested?
Pass all test cases.
Author: Dongjoon Hyun <dongjoon@apache.org>
Closes#20610 from dongjoon-hyun/SPARK-ORC-DISABLE.
## What changes were proposed in this pull request?
This PR proposes to add an alias 'names' of 'fieldNames' in Scala. Please see the discussion in [SPARK-20090](https://issues.apache.org/jira/browse/SPARK-20090).
## How was this patch tested?
Unit tests added in `DataTypeSuite.scala`.
Author: hyukjinkwon <gurwls223@gmail.com>
Closes#20545 from HyukjinKwon/SPARK-23359.
## What changes were proposed in this pull request?
This PR is to revert the PR https://github.com/apache/spark/pull/20302, because it causes a regression.
## How was this patch tested?
N/A
Author: gatorsmile <gatorsmile@gmail.com>
Closes#20614 from gatorsmile/revertJsonFix.
## What changes were proposed in this pull request?
In this upcoming 2.3 release, we changed the interface of `ScalaUDF`. Unfortunately, some Spark packages (e.g., spark-deep-learning) are using our internal class `ScalaUDF`. In the release 2.3, we added new parameters into this class. The users hit the binary compatibility issues and got the exception:
```
> java.lang.NoSuchMethodError: org.apache.spark.sql.catalyst.expressions.ScalaUDF.<init>(Ljava/lang/Object;Lorg/apache/spark/sql/types/DataType;Lscala/collection/Seq;Lscala/collection/Seq;Lscala/Option;)V
```
This PR is to improve the backward compatibility. However, we definitely should not encourage the external packages to use our internal classes. This might make us hard to maintain/develop the codes in Spark.
## How was this patch tested?
N/A
Author: gatorsmile <gatorsmile@gmail.com>
Closes#20591 from gatorsmile/scalaUDF.
## What changes were proposed in this pull request?
Added flag ignoreNullability to DataType.equalsStructurally.
The previous semantic is for ignoreNullability=false.
When ignoreNullability=true equalsStructurally ignores nullability of contained types (map key types, value types, array element types, structure field types).
In.checkInputTypes calls equalsStructurally to check if the children types match. They should match regardless of nullability (which is just a hint), so it is now called with ignoreNullability=true.
## How was this patch tested?
New test in SubquerySuite
Author: Bogdan Raducanu <bogdan@databricks.com>
Closes#20548 from bogdanrdc/SPARK-23316.
## What changes were proposed in this pull request?
This PR targets to explicitly specify supported types in Pandas UDFs.
The main change here is to add a deduplicated and explicit type checking in `returnType` ahead with documenting this; however, it happened to fix multiple things.
1. Currently, we don't support `BinaryType` in Pandas UDFs, for example, see:
```python
from pyspark.sql.functions import pandas_udf
pudf = pandas_udf(lambda x: x, "binary")
df = spark.createDataFrame([[bytearray(1)]])
df.select(pudf("_1")).show()
```
```
...
TypeError: Unsupported type in conversion to Arrow: BinaryType
```
We can document this behaviour for its guide.
2. Also, the grouped aggregate Pandas UDF fails fast on `ArrayType` but seems we can support this case.
```python
from pyspark.sql.functions import pandas_udf, PandasUDFType
foo = pandas_udf(lambda v: v.mean(), 'array<double>', PandasUDFType.GROUPED_AGG)
df = spark.range(100).selectExpr("id", "array(id) as value")
df.groupBy("id").agg(foo("value")).show()
```
```
...
NotImplementedError: ArrayType, StructType and MapType are not supported with PandasUDFType.GROUPED_AGG
```
3. Since we can check the return type ahead, we can fail fast before actual execution.
```python
# we can fail fast at this stage because we know the schema ahead
pandas_udf(lambda x: x, BinaryType())
```
## How was this patch tested?
Manually tested and unit tests for `BinaryType` and `ArrayType(...)` were added.
Author: hyukjinkwon <gurwls223@gmail.com>
Closes#20531 from HyukjinKwon/pudf-cleanup.
## What changes were proposed in this pull request?
Update the description and tests of three external API or functions `createFunction `, `length` and `repartitionByRange `
## How was this patch tested?
N/A
Author: gatorsmile <gatorsmile@gmail.com>
Closes#20495 from gatorsmile/updateFunc.
## What changes were proposed in this pull request?
In the current test case for CombineTypedFilters, we lack the test of FilterFunction, so let's add it.
In addition, in TypedFilterOptimizationSuite's existing test cases, Let's extract a common LocalRelation.
## How was this patch tested?
add new test cases.
Author: caoxuewen <cao.xuewen@zte.com.cn>
Closes#20482 from heary-cao/TypedFilterOptimizationSuite.
## What changes were proposed in this pull request?
https://issues.apache.org/jira/browse/SPARK-23309 reported a performance regression about cached table in Spark 2.3. While the investigating is still going on, this PR adds a conf to turn off the vectorized cache reader, to unblock the 2.3 release.
## How was this patch tested?
a new test
Author: Wenchen Fan <wenchen@databricks.com>
Closes#20483 from cloud-fan/cache.
## What changes were proposed in this pull request?
SpecifiedWindowFrame.defaultWindowFrame(hasOrderSpecification, acceptWindowFrame) was designed to handle the cases when some Window functions don't support setting a window frame (e.g. rank). However this param is never used.
We may inline the whole of this function to simplify the code.
## How was this patch tested?
Existing tests.
Author: Xingbo Jiang <xingbo.jiang@databricks.com>
Closes#20463 from jiangxb1987/defaultWindowFrame.
## What changes were proposed in this pull request?
This PR include the following changes:
- Make the capacity of `VectorizedParquetRecordReader` configurable;
- Make the capacity of `OrcColumnarBatchReader` configurable;
- Update the error message when required capacity in writable columnar vector cannot be fulfilled.
## How was this patch tested?
N/A
Author: Xingbo Jiang <xingbo.jiang@databricks.com>
Closes#20361 from jiangxb1987/vectorCapacity.
## What changes were proposed in this pull request?
Here is the test snippet.
``` SQL
scala> Seq[(Integer, Integer)](
| (1, 1),
| (1, 3),
| (2, 3),
| (3, 3),
| (4, null),
| (5, null)
| ).toDF("key", "value").createOrReplaceTempView("src")
scala> sql(
| """
| |SELECT MAX(value) as value, key as col2
| |FROM src
| |GROUP BY key
| |ORDER BY value desc, key
| """.stripMargin).show
+-----+----+
|value|col2|
+-----+----+
| 3| 3|
| 3| 2|
| 3| 1|
| null| 5|
| null| 4|
+-----+----+
```SQL
Here is the explain output :
```SQL
== Parsed Logical Plan ==
'Sort ['value DESC NULLS LAST, 'key ASC NULLS FIRST], true
+- 'Aggregate ['key], ['MAX('value) AS value#9, 'key AS col2#10]
+- 'UnresolvedRelation `src`
== Analyzed Logical Plan ==
value: int, col2: int
Project [value#9, col2#10]
+- Sort [value#9 DESC NULLS LAST, col2#10 DESC NULLS LAST], true
+- Aggregate [key#5], [max(value#6) AS value#9, key#5 AS col2#10]
+- SubqueryAlias src
+- Project [_1#2 AS key#5, _2#3 AS value#6]
+- LocalRelation [_1#2, _2#3]
``` SQL
The sort direction is being wrongly changed from ASC to DSC while resolving ```Sort``` in
resolveAggregateFunctions.
The above testcase models TPCDS-Q71 and thus we have the same issue in Q71 as well.
## How was this patch tested?
A few tests are added in SQLQuerySuite.
Author: Dilip Biswal <dbiswal@us.ibm.com>
Closes#20453 from dilipbiswal/local_spark.
## What changes were proposed in this pull request?
This PR is to fix the `ReplaceExceptWithFilter` rule when the right's Filter contains the references that are not in the left output.
Before this PR, we got the error like
```
java.util.NoSuchElementException: key not found: a
at scala.collection.MapLike$class.default(MapLike.scala:228)
at scala.collection.AbstractMap.default(Map.scala:59)
at scala.collection.MapLike$class.apply(MapLike.scala:141)
at scala.collection.AbstractMap.apply(Map.scala:59)
```
After this PR, `ReplaceExceptWithFilter ` will not take an effect in this case.
## How was this patch tested?
Added tests
Author: gatorsmile <gatorsmile@gmail.com>
Closes#20444 from gatorsmile/fixReplaceExceptWithFilter.
## What changes were proposed in this pull request?
Still saw the performance regression introduced by `spark.sql.codegen.hugeMethodLimit` in our internal workloads. There are two major issues in the current solution.
- The size of the complied byte code is not identical to the bytecode size of the method. The detection is still not accurate.
- The bytecode size of a single operator (e.g., `SerializeFromObject`) could still exceed 8K limit. We saw the performance regression in such scenario.
Since it is close to the release of 2.3, we decide to increase it to 64K for avoiding the perf regression.
## How was this patch tested?
N/A
Author: gatorsmile <gatorsmile@gmail.com>
Closes#20434 from gatorsmile/revertConf.
## What changes were proposed in this pull request?
Rename the public APIs and names of pandas udfs.
- `PANDAS SCALAR UDF` -> `SCALAR PANDAS UDF`
- `PANDAS GROUP MAP UDF` -> `GROUPED MAP PANDAS UDF`
- `PANDAS GROUP AGG UDF` -> `GROUPED AGG PANDAS UDF`
## How was this patch tested?
The existing tests
Author: gatorsmile <gatorsmile@gmail.com>
Closes#20428 from gatorsmile/renamePandasUDFs.
## What changes were proposed in this pull request?
This is a followup to #19575 which added a section on setting max Arrow record batches and this will externalize the conf that was referenced in the docs.
## How was this patch tested?
NA
Author: Bryan Cutler <cutlerb@gmail.com>
Closes#20423 from BryanCutler/arrow-user-doc-externalize-maxRecordsPerBatch-SPARK-22221.
## What changes were proposed in this pull request?
It is a common pattern to apply multiple transforms to a `Dataset` (using `Dataset.withColumn` for example. This is currently quite expensive because we run `CheckAnalysis` on the full plan and create an encoder for each intermediate `Dataset`.
This PR extends the usage of the `AnalysisBarrier` to include `CheckAnalysis`. By doing this we hide the already analyzed plan from `CheckAnalysis` because barrier is a `LeafNode`. The `AnalysisBarrier` is in the `FinishAnalysis` phase of the optimizer.
We also make binding the `Dataset` encoder lazy. The bound encoder is only needed when we materialize the dataset.
## How was this patch tested?
Existing test should cover this.
Author: Herman van Hovell <hvanhovell@databricks.com>
Closes#20402 from hvanhovell/SPARK-23223.
## What changes were proposed in this pull request?
Currently, all Aggregate operations will go into RemoveRepetitionFromGroupExpressions, but there is no group expression or there is no duplicate group expression in group expression, we not need copy for logic plan.
## How was this patch tested?
the existed test case.
Author: caoxuewen <cao.xuewen@zte.com.cn>
Closes#20375 from heary-cao/RepetitionGroupExpressions.
## What changes were proposed in this pull request?
Currently we have `ReadTask` in data source v2 reader, while in writer we have `DataWriterFactory`.
To make the naming consistent and better, renaming `ReadTask` to `DataReaderFactory`.
## How was this patch tested?
Unit test
Author: Wang Gengliang <ltnwgl@gmail.com>
Closes#20397 from gengliangwang/rename.
## What changes were proposed in this pull request?
This PR proposes to expose few internal configurations found in the documentation.
Also it fixes the description for `spark.sql.execution.arrow.enabled`.
It's quite self-explanatory.
## How was this patch tested?
N/A
Author: hyukjinkwon <gurwls223@gmail.com>
Closes#20403 from HyukjinKwon/minor-doc-arrow.
## What changes were proposed in this pull request?
Replace streaming V2 sinks with a unified StreamWriteSupport interface, with a shim to use it with microbatch execution.
Add a new SQL config to use for disabling V2 sinks, falling back to the V1 sink implementation.
## How was this patch tested?
Existing tests, which in the case of Kafka (the only existing continuous V2 sink) now use V2 for microbatch.
Author: Jose Torres <jose@databricks.com>
Closes#20369 from jose-torres/streaming-sink.
## What changes were proposed in this pull request?
Currently shuffle repartition uses RoundRobinPartitioning, the generated result is nondeterministic since the sequence of input rows are not determined.
The bug can be triggered when there is a repartition call following a shuffle (which would lead to non-deterministic row ordering), as the pattern shows below:
upstream stage -> repartition stage -> result stage
(-> indicate a shuffle)
When one of the executors process goes down, some tasks on the repartition stage will be retried and generate inconsistent ordering, and some tasks of the result stage will be retried generating different data.
The following code returns 931532, instead of 1000000:
```
import scala.sys.process._
import org.apache.spark.TaskContext
val res = spark.range(0, 1000 * 1000, 1).repartition(200).map { x =>
x
}.repartition(200).map { x =>
if (TaskContext.get.attemptNumber == 0 && TaskContext.get.partitionId < 2) {
throw new Exception("pkill -f java".!!)
}
x
}
res.distinct().count()
```
In this PR, we propose a most straight-forward way to fix this problem by performing a local sort before partitioning, after we make the input row ordering deterministic, the function from rows to partitions is fully deterministic too.
The downside of the approach is that with extra local sort inserted, the performance of repartition() will go down, so we add a new config named `spark.sql.execution.sortBeforeRepartition` to control whether this patch is applied. The patch is default enabled to be safe-by-default, but user may choose to manually turn it off to avoid performance regression.
This patch also changes the output rows ordering of repartition(), that leads to a bunch of test cases failure because they are comparing the results directly.
## How was this patch tested?
Add unit test in ExchangeSuite.
With this patch(and `spark.sql.execution.sortBeforeRepartition` set to true), the following query returns 1000000:
```
import scala.sys.process._
import org.apache.spark.TaskContext
spark.conf.set("spark.sql.execution.sortBeforeRepartition", "true")
val res = spark.range(0, 1000 * 1000, 1).repartition(200).map { x =>
x
}.repartition(200).map { x =>
if (TaskContext.get.attemptNumber == 0 && TaskContext.get.partitionId < 2) {
throw new Exception("pkill -f java".!!)
}
x
}
res.distinct().count()
res7: Long = 1000000
```
Author: Xingbo Jiang <xingbo.jiang@databricks.com>
Closes#20393 from jiangxb1987/shuffle-repartition.
## What changes were proposed in this pull request?
**Proposal**
Add a per-query ID to the codegen stages as represented by `WholeStageCodegenExec` operators. This ID will be used in
- the explain output of the physical plan, and in
- the generated class name.
Specifically, this ID will be stable within a query, counting up from 1 in depth-first post-order for all the `WholeStageCodegenExec` inserted into a plan.
The ID value 0 is reserved for "free-floating" `WholeStageCodegenExec` objects, which may have been created for one-off purposes, e.g. for fallback handling of codegen stages that failed to codegen the whole stage and wishes to codegen a subset of the children operators (as seen in `org.apache.spark.sql.execution.FileSourceScanExec#doExecute`).
Example: for the following query:
```scala
scala> spark.conf.set("spark.sql.autoBroadcastJoinThreshold", 1)
scala> val df1 = spark.range(10).select('id as 'x, 'id + 1 as 'y).orderBy('x).select('x + 1 as 'z, 'y)
df1: org.apache.spark.sql.DataFrame = [z: bigint, y: bigint]
scala> val df2 = spark.range(5)
df2: org.apache.spark.sql.Dataset[Long] = [id: bigint]
scala> val query = df1.join(df2, 'z === 'id)
query: org.apache.spark.sql.DataFrame = [z: bigint, y: bigint ... 1 more field]
```
The explain output before the change is:
```scala
scala> query.explain
== Physical Plan ==
*SortMergeJoin [z#9L], [id#13L], Inner
:- *Sort [z#9L ASC NULLS FIRST], false, 0
: +- Exchange hashpartitioning(z#9L, 200)
: +- *Project [(x#3L + 1) AS z#9L, y#4L]
: +- *Sort [x#3L ASC NULLS FIRST], true, 0
: +- Exchange rangepartitioning(x#3L ASC NULLS FIRST, 200)
: +- *Project [id#0L AS x#3L, (id#0L + 1) AS y#4L]
: +- *Range (0, 10, step=1, splits=8)
+- *Sort [id#13L ASC NULLS FIRST], false, 0
+- Exchange hashpartitioning(id#13L, 200)
+- *Range (0, 5, step=1, splits=8)
```
Note how codegen'd operators are annotated with a prefix `"*"`. See how the `SortMergeJoin` operator and its direct children `Sort` operators are adjacent and all annotated with the `"*"`, so it's hard to tell they're actually in separate codegen stages.
and after this change it'll be:
```scala
scala> query.explain
== Physical Plan ==
*(6) SortMergeJoin [z#9L], [id#13L], Inner
:- *(3) Sort [z#9L ASC NULLS FIRST], false, 0
: +- Exchange hashpartitioning(z#9L, 200)
: +- *(2) Project [(x#3L + 1) AS z#9L, y#4L]
: +- *(2) Sort [x#3L ASC NULLS FIRST], true, 0
: +- Exchange rangepartitioning(x#3L ASC NULLS FIRST, 200)
: +- *(1) Project [id#0L AS x#3L, (id#0L + 1) AS y#4L]
: +- *(1) Range (0, 10, step=1, splits=8)
+- *(5) Sort [id#13L ASC NULLS FIRST], false, 0
+- Exchange hashpartitioning(id#13L, 200)
+- *(4) Range (0, 5, step=1, splits=8)
```
Note that the annotated prefix becomes `"*(id) "`. See how the `SortMergeJoin` operator and its direct children `Sort` operators have different codegen stage IDs.
It'll also show up in the name of the generated class, as a suffix in the format of `GeneratedClass$GeneratedIterator$id`.
For example, note how `GeneratedClass$GeneratedIteratorForCodegenStage3` and `GeneratedClass$GeneratedIteratorForCodegenStage6` in the following stack trace corresponds to the IDs shown in the explain output above:
```
"Executor task launch worker for task 42412957" daemon prio=5 tid=0x58 nid=NA runnable
java.lang.Thread.State: RUNNABLE
at org.apache.spark.sql.execution.UnsafeExternalRowSorter.insertRow(UnsafeExternalRowSorter.java:109)
at org.apache.spark.sql.catalyst.expressions.GeneratedClass$GeneratedIteratorForCodegenStage3.sort_addToSorter$(generated.java:32)
at org.apache.spark.sql.catalyst.expressions.GeneratedClass$GeneratedIteratorForCodegenStage3.processNext(generated.java:41)
at org.apache.spark.sql.execution.BufferedRowIterator.hasNext(BufferedRowIterator.java:43)
at org.apache.spark.sql.execution.WholeStageCodegenExec$$anonfun$9$$anon$1.hasNext(WholeStageCodegenExec.scala:494)
at org.apache.spark.sql.catalyst.expressions.GeneratedClass$GeneratedIteratorForCodegenStage6.findNextInnerJoinRows$(generated.java:42)
at org.apache.spark.sql.catalyst.expressions.GeneratedClass$GeneratedIteratorForCodegenStage6.processNext(generated.java:101)
at org.apache.spark.sql.execution.BufferedRowIterator.hasNext(BufferedRowIterator.java:43)
at org.apache.spark.sql.execution.WholeStageCodegenExec$$anonfun$11$$anon$2.hasNext(WholeStageCodegenExec.scala:513)
at org.apache.spark.sql.execution.SparkPlan$$anonfun$2.apply(SparkPlan.scala:253)
at org.apache.spark.sql.execution.SparkPlan$$anonfun$2.apply(SparkPlan.scala:247)
at org.apache.spark.rdd.RDD$$anonfun$mapPartitionsInternal$1$$anonfun$apply$25.apply(RDD.scala:828)
at org.apache.spark.rdd.RDD$$anonfun$mapPartitionsInternal$1$$anonfun$apply$25.apply(RDD.scala:828)
at org.apache.spark.rdd.MapPartitionsRDD.compute(MapPartitionsRDD.scala:38)
at org.apache.spark.rdd.RDD.computeOrReadCheckpoint(RDD.scala:324)
at org.apache.spark.rdd.RDD.iterator(RDD.scala:288)
at org.apache.spark.scheduler.ResultTask.runTask(ResultTask.scala:87)
at org.apache.spark.scheduler.Task.run(Task.scala:109)
at org.apache.spark.executor.Executor$TaskRunner.run(Executor.scala:345)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
at java.lang.Thread.run(Thread.java:748)
```
**Rationale**
Right now, the codegen from Spark SQL lacks the means to differentiate between a couple of things:
1. It's hard to tell which physical operators are in the same WholeStageCodegen stage. Note that this "stage" is a separate notion from Spark's RDD execution stages; this one is only to delineate codegen units.
There can be adjacent physical operators that are both codegen'd but are in separate codegen stages. Some of this is due to hacky implementation details, such as the case with `SortMergeJoin` and its `Sort` inputs -- they're hard coded to be split into separate stages although both are codegen'd.
When printing out the explain output of the physical plan, you'd only see the codegen'd physical operators annotated with a preceding star (`'*'`) but would have no way to figure out if they're in the same stage.
2. Performance/error diagnosis
The generated code has class/method names that are hard to differentiate between queries or even between codegen stages within the same query. If we use a Java-level profiler to collect profiles, or if we encounter a Java-level exception with a stack trace in it, it's really hard to tell which part of a query it's at.
By introducing a per-query codegen stage ID, we'd at least be able to know which codegen stage (and in turn, which group of physical operators) was a profile tick or an exception happened.
The reason why this proposal uses a per-query ID is because it's stable within a query, so that multiple runs of the same query will see the same resulting IDs. This both benefits understandability for users, and also it plays well with the codegen cache in Spark SQL which uses the generated source code as the key.
The downside to using per-query IDs as opposed to a per-session or globally incrementing ID is of course we can't tell apart different query runs with this ID alone. But for now I believe this is a good enough tradeoff.
## How was this patch tested?
Existing tests. This PR does not involve any runtime behavior changes other than some name changes.
The SQL query test suites that compares explain outputs have been updates to ignore the newly added `codegenStageId`.
Author: Kris Mok <kris.mok@databricks.com>
Closes#20224 from rednaxelafx/wsc-codegenstageid.
## What changes were proposed in this pull request?
It has been observed in SPARK-21603 that whole-stage codegen suffers performance degradation, if the generated functions are too long to be optimized by JIT.
We basically produce a single function to incorporate generated codes from all physical operators in whole-stage. Thus, it is possibly to grow the size of generated function over a threshold that we can't have JIT optimization for it anymore.
This patch is trying to decouple the logic of consuming rows in physical operators to avoid a giant function processing rows.
## How was this patch tested?
Added tests.
Author: Liang-Chi Hsieh <viirya@gmail.com>
Closes#18931 from viirya/SPARK-21717.
## What changes were proposed in this pull request?
The `GenArrayData.genCodeToCreateArrayData` produces illegal java code when code splitting is enabled. This is used in `CreateArray` and `CreateMap` expressions for complex object arrays.
This issue is caused by a typo.
## How was this patch tested?
Added a regression test in `complexTypesSuite`.
Author: Herman van Hovell <hvanhovell@databricks.com>
Closes#20391 from hvanhovell/SPARK-23208.
## What changes were proposed in this pull request?
This PR is repaired as follows
1、update y -> x in "left outer join" test case ,maybe is mistake.
2、add a new test case："left outer join and left sides are limited"
3、add a new test case："left outer join and right sides are limited"
4、add a new test case: "right outer join and right sides are limited"
5、add a new test case: "right outer join and left sides are limited"
6、Remove annotations without code implementation
## How was this patch tested?
add new unit test case.
Author: caoxuewen <cao.xuewen@zte.com.cn>
Closes#20381 from heary-cao/LimitPushdownSuite.
## What changes were proposed in this pull request?
We need to override the prettyName for bit_length and octet_length for getting the expected auto-generated alias name.
## How was this patch tested?
The existing tests
Author: gatorsmile <gatorsmile@gmail.com>
Closes#20358 from gatorsmile/test2.3More.
## What changes were proposed in this pull request?
Add support for using pandas UDFs with groupby().agg().
This PR introduces a new type of pandas UDF - group aggregate pandas UDF. This type of UDF defines a transformation of multiple pandas Series -> a scalar value. Group aggregate pandas UDFs can be used with groupby().agg(). Note group aggregate pandas UDF doesn't support partial aggregation, i.e., a full shuffle is required.
This PR doesn't support group aggregate pandas UDFs that return ArrayType, StructType or MapType. Support for these types is left for future PR.
## How was this patch tested?
GroupbyAggPandasUDFTests
Author: Li Jin <ice.xelloss@gmail.com>
Closes#19872 from icexelloss/SPARK-22274-groupby-agg.
## What changes were proposed in this pull request?
a new interface which allows data source to report partitioning and avoid shuffle at Spark side.
The design is pretty like the internal distribution/partitioing framework. Spark defines a `Distribution` interfaces and several concrete implementations, and ask the data source to report a `Partitioning`, the `Partitioning` should tell Spark if it can satisfy a `Distribution` or not.
## How was this patch tested?
new test
Author: Wenchen Fan <wenchen@databricks.com>
Closes#20201 from cloud-fan/partition-reporting.
## What changes were proposed in this pull request?
Typo fixes
## How was this patch tested?
Local build / Doc-only changes
Author: Jacek Laskowski <jacek@japila.pl>
Closes#20344 from jaceklaskowski/typo-fixes.
## What changes were proposed in this pull request?
Several improvements:
* provide a default implementation for the batch get methods
* rename `getChildColumn` to `getChild`, which is more concise
* remove `getStruct(int, int)`, it's only used to simplify the codegen, which is an internal thing, we should not add a public API for this purpose.
## How was this patch tested?
existing tests
Author: Wenchen Fan <wenchen@databricks.com>
Closes#20277 from cloud-fan/column-vector.
## What changes were proposed in this pull request?
CheckCartesianProduct raises an AnalysisException also when the join condition is always false/null. In this case, we shouldn't raise it, since the result will not be a cartesian product.
## How was this patch tested?
added UT
Author: Marco Gaido <marcogaido91@gmail.com>
Closes#20333 from mgaido91/SPARK-23087.
## What changes were proposed in this pull request?
This is a follow-up of #20246.
If a UDT in Python doesn't have its corresponding Scala UDT, cast to string will be the raw string of the internal value, e.g. `"org.apache.spark.sql.catalyst.expressions.UnsafeArrayDataxxxxxxxx"` if the internal type is `ArrayType`.
This pr fixes it by using its `sqlType` casting.
## How was this patch tested?
Added a test and existing tests.
Author: Takuya UESHIN <ueshin@databricks.com>
Closes#20306 from ueshin/issues/SPARK-23054/fup1.
## What changes were proposed in this pull request?
There were two related fixes regarding `from_json`, `get_json_object` and `json_tuple` ([Fix#1](c8803c0685),
[Fix#2](86174ea89b)), but they weren't comprehensive it seems. I wanted to extend those fixes to all the parsers, and add tests for each case.
## How was this patch tested?
Regression tests
Author: Burak Yavuz <brkyvz@gmail.com>
Closes#20302 from brkyvz/json-invfix.
## What changes were proposed in this pull request?
When there is an operation between Decimals and the result is a number which is not representable exactly with the result's precision and scale, Spark is returning `NULL`. This was done to reflect Hive's behavior, but it is against SQL ANSI 2011, which states that "If the result cannot be represented exactly in the result type, then whether it is rounded or truncated is implementation-defined". Moreover, Hive now changed its behavior in order to respect the standard, thanks to HIVE-15331.
Therefore, the PR propose to:
- update the rules to determine the result precision and scale according to the new Hive's ones introduces in HIVE-15331;
- round the result of the operations, when it is not representable exactly with the result's precision and scale, instead of returning `NULL`
- introduce a new config `spark.sql.decimalOperations.allowPrecisionLoss` which default to `true` (ie. the new behavior) in order to allow users to switch back to the previous one.
Hive behavior reflects SQLServer's one. The only difference is that the precision and scale are adjusted for all the arithmetic operations in Hive, while SQL Server is said to do so only for multiplications and divisions in the documentation. This PR follows Hive's behavior.
A more detailed explanation is available here: https://mail-archives.apache.org/mod_mbox/spark-dev/201712.mbox/%3CCAEorWNAJ4TxJR9NBcgSFMD_VxTg8qVxusjP%2BAJP-x%2BJV9zH-yA%40mail.gmail.com%3E.
## How was this patch tested?
modified and added UTs. Comparisons with results of Hive and SQLServer.
Author: Marco Gaido <marcogaido91@gmail.com>
Closes#20023 from mgaido91/SPARK-22036.
## What changes were proposed in this pull request?
Previously, PR #19201 fix the problem of non-converging constraints.
After that PR #19149 improve the loop and constraints is inferred only once.
So the problem of non-converging constraints is gone.
However, the case below will fail.
```
spark.range(5).write.saveAsTable("t")
val t = spark.read.table("t")
val left = t.withColumn("xid", $"id" + lit(1)).as("x")
val right = t.withColumnRenamed("id", "xid").as("y")
val df = left.join(right, "xid").filter("id = 3").toDF()
checkAnswer(df, Row(4, 3))
```
Because `aliasMap` replace all the aliased child. See the test case in PR for details.
This PR is to fix this bug by removing useless code for preventing non-converging constraints.
It can be also fixed with #20270, but this is much simpler and clean up the code.
## How was this patch tested?
Unit test
Author: Wang Gengliang <ltnwgl@gmail.com>
Closes#20278 from gengliangwang/FixConstraintSimple.
## What changes were proposed in this pull request?
ORC filter push-down is disabled by default from the beginning, [SPARK-2883](aa31e431fc (diff-41ef65b9ef5b518f77e2a03559893f4dR149)
).
Now, Apache Spark starts to depend on Apache ORC 1.4.1. For Apache Spark 2.3, this PR turns on ORC filter push-down by default like Parquet ([SPARK-9207](https://issues.apache.org/jira/browse/SPARK-21783)) as a part of [SPARK-20901](https://issues.apache.org/jira/browse/SPARK-20901), "Feature parity for ORC with Parquet".
## How was this patch tested?
Pass the existing tests.
Author: Dongjoon Hyun <dongjoon@apache.org>
Closes#20265 from dongjoon-hyun/SPARK-21783.
## What changes were proposed in this pull request?
There are already quite a few integration tests using window frames, but the unit tests coverage is not ideal.
In this PR the already existing tests are reorganized, extended and where gaps found additional cases added.
## How was this patch tested?
Automated: Pass the Jenkins.
Author: Gabor Somogyi <gabor.g.somogyi@gmail.com>
Closes#20019 from gaborgsomogyi/SPARK-22361.
## What changes were proposed in this pull request?
The following SQL involving scalar correlated query returns a map exception.
``` SQL
SELECT t1a
FROM t1
WHERE t1a = (SELECT count(*)
FROM t2
WHERE t2c = t1c
HAVING count(*) >= 1)
```
``` SQL
key not found: ExprId(278,786682bb-41f9-4bd5-a397-928272cc8e4e)
java.util.NoSuchElementException: key not found: ExprId(278,786682bb-41f9-4bd5-a397-928272cc8e4e)
at scala.collection.MapLike$class.default(MapLike.scala:228)
at scala.collection.AbstractMap.default(Map.scala:59)
at scala.collection.MapLike$class.apply(MapLike.scala:141)
at scala.collection.AbstractMap.apply(Map.scala:59)
at org.apache.spark.sql.catalyst.optimizer.RewriteCorrelatedScalarSubquery$.org$apache$spark$sql$catalyst$optimizer$RewriteCorrelatedScalarSubquery$$evalSubqueryOnZeroTups(subquery.scala:378)
at org.apache.spark.sql.catalyst.optimizer.RewriteCorrelatedScalarSubquery$$anonfun$org$apache$spark$sql$catalyst$optimizer$RewriteCorrelatedScalarSubquery$$constructLeftJoins$1.apply(subquery.scala:430)
at org.apache.spark.sql.catalyst.optimizer.RewriteCorrelatedScalarSubquery$$anonfun$org$apache$spark$sql$catalyst$optimizer$RewriteCorrelatedScalarSubquery$$constructLeftJoins$1.apply(subquery.scala:426)
```
In this case, after evaluating the HAVING clause "count(*) > 1" statically
against the binding of aggregtation result on empty input, we determine
that this query will not have a the count bug. We should simply return
the evalSubqueryOnZeroTups with empty value.
(Please fill in changes proposed in this fix)
## How was this patch tested?
A new test was added in the Subquery bucket.
Author: Dilip Biswal <dbiswal@us.ibm.com>
Closes#20283 from dilipbiswal/scalar-count-defect.
## What changes were proposed in this pull request?
When a user puts the wrong number of parameters in a function, an AnalysisException is thrown. If the function is a UDF, he user is told how many parameters the function expected and how many he/she put. If the function, instead, is a built-in one, no information about the number of parameters expected and the actual one is provided. This can help in some cases, to debug the errors (eg. bad quotes escaping may lead to a different number of parameters than expected, etc. etc.)
The PR adds the information about the number of parameters passed and the expected one, analogously to what happens for UDF.
## How was this patch tested?
modified existing UT + manual test
Author: Marco Gaido <marcogaido91@gmail.com>
Closes#20271 from mgaido91/SPARK-23080.
## What changes were proposed in this pull request?
Problem: it throw TempTableAlreadyExistsException and output "Temporary table '$table' already exists" when we create temp view by using org.apache.spark.sql.catalyst.catalog.GlobalTempViewManager#create, it's improper.
So fix improper information about TempTableAlreadyExistsException when create temp view:
change "Temporary table" to "Temporary view"
## How was this patch tested?
test("rename temporary view - destination table already exists, with: CREATE TEMPORARY view")
test("rename temporary view - destination table with database name,with:CREATE TEMPORARY view")
Author: xubo245 <601450868@qq.com>
Closes#20227 from xubo245/fixDeprecated.
## What changes were proposed in this pull request?
Remove `MaxPermSize` for `sql` module
## How was this patch tested?
Manually tested.
Author: Yuming Wang <yumwang@ebay.com>
Closes#20268 from wangyum/SPARK-19550-MaxPermSize.
## What changes were proposed in this pull request?
This pr fixed the issue when casting `UserDefinedType`s into strings;
```
>>> from pyspark.ml.classification import MultilayerPerceptronClassifier
>>> from pyspark.ml.linalg import Vectors
>>> df = spark.createDataFrame([(0.0, Vectors.dense([0.0, 0.0])), (1.0, Vectors.dense([0.0, 1.0]))], ["label", "features"])
>>> df.selectExpr("CAST(features AS STRING)").show(truncate = False)
+-------------------------------------------+
|features |
+-------------------------------------------+
|[6,1,0,0,2800000020,2,0,0,0] |
|[6,1,0,0,2800000020,2,0,0,3ff0000000000000]|
+-------------------------------------------+
```
The root cause is that `Cast` handles input data as `UserDefinedType.sqlType`(this is underlying storage type), so we should pass data into `UserDefinedType.deserialize` then `toString`.
This pr modified the result into;
```
+---------+
|features |
+---------+
|[0.0,0.0]|
|[0.0,1.0]|
+---------+
```
## How was this patch tested?
Added tests in `UserDefinedTypeSuite `.
Author: Takeshi Yamamuro <yamamuro@apache.org>
Closes#20246 from maropu/SPARK-23054.
## What changes were proposed in this pull request?
SHOW DATABASES (LIKE pattern = STRING)? Can be like the back increase?
When using this command, LIKE keyword can be removed.
You can refer to the SHOW TABLES command, SHOW TABLES 'test *' and SHOW TABELS like 'test *' can be used.
Similarly SHOW DATABASES 'test *' and SHOW DATABASES like 'test *' can be used.
## How was this patch tested?
unit tests manual tests
Please review http://spark.apache.org/contributing.html before opening a pull request.
Author: guoxiaolong <guo.xiaolong1@zte.com.cn>
Closes#20194 from guoxiaolongzte/SPARK-22999.
## What changes were proposed in this pull request?
`AnalysisBarrier` in the current master cuts off explain results for parsed logical plans;
```
scala> Seq((1, 1)).toDF("a", "b").groupBy("a").count().sample(0.1).explain(true)
== Parsed Logical Plan ==
Sample 0.0, 0.1, false, -7661439431999668039
+- AnalysisBarrier Aggregate [a#5], [a#5, count(1) AS count#14L]
```
To fix this, `AnalysisBarrier` needs to override `innerChildren` and this pr changed the output to;
```
== Parsed Logical Plan ==
Sample 0.0, 0.1, false, -5086223488015741426
+- AnalysisBarrier
+- Aggregate [a#5], [a#5, count(1) AS count#14L]
+- Project [_1#2 AS a#5, _2#3 AS b#6]
+- LocalRelation [_1#2, _2#3]
```
## How was this patch tested?
Added tests in `DataFrameSuite`.
Author: Takeshi Yamamuro <yamamuro@apache.org>
Closes#20247 from maropu/SPARK-23021-2.
## What changes were proposed in this pull request?
as per discussion in https://github.com/apache/spark/pull/19864#discussion_r156847927
the current HadoopFsRelation is purely based on the underlying file size which is not accurate and makes the execution vulnerable to errors like OOM
Users can enable CBO with the functionalities in https://github.com/apache/spark/pull/19864 to avoid this issue
This JIRA proposes to add a configurable factor to sizeInBytes method in HadoopFsRelation class so that users can mitigate this problem without CBO
## How was this patch tested?
Existing tests
Author: CodingCat <zhunansjtu@gmail.com>
Author: Nan Zhu <nanzhu@uber.com>
Closes#20072 from CodingCat/SPARK-22790.
## What changes were proposed in this pull request?
This patch bumps the master branch version to `2.4.0-SNAPSHOT`.
## How was this patch tested?
N/A
Author: gatorsmile <gatorsmile@gmail.com>
Closes#20222 from gatorsmile/bump24.
## What changes were proposed in this pull request?
Add support for `Null` type in the `schemaFor` method for Scala reflection.
## How was this patch tested?
Added UT
Author: Marco Gaido <marcogaido91@gmail.com>
Closes#20219 from mgaido91/SPARK-23025.
## What changes were proposed in this pull request?
(courtesy of liancheng)
Spark SQL supports both global aggregation and grouping aggregation. Global aggregation always return a single row with the initial aggregation state as the output, even there are zero input rows. Spark implements this by simply checking the number of grouping keys and treats an aggregation as a global aggregation if it has zero grouping keys.
However, this simple principle drops the ball in the following case:
```scala
spark.emptyDataFrame.dropDuplicates().agg(count($"*") as "c").show()
// +---+
// | c |
// +---+
// | 1 |
// +---+
```
The reason is that:
1. `df.dropDuplicates()` is roughly translated into something equivalent to:
```scala
val allColumns = df.columns.map { col }
df.groupBy(allColumns: _*).agg(allColumns.head, allColumns.tail: _*)
```
This translation is implemented in the rule `ReplaceDeduplicateWithAggregate`.
2. `spark.emptyDataFrame` contains zero columns and zero rows.
Therefore, rule `ReplaceDeduplicateWithAggregate` makes a confusing transformation roughly equivalent to the following one:
```scala
spark.emptyDataFrame.dropDuplicates()
=> spark.emptyDataFrame.groupBy().agg(Map.empty[String, String])
```
The above transformation is confusing because the resulting aggregate operator contains no grouping keys (because `emptyDataFrame` contains no columns), and gets recognized as a global aggregation. As a result, Spark SQL allocates a single row filled by the initial aggregation state and uses it as the output, and returns a wrong result.
To fix this issue, this PR tweaks `ReplaceDeduplicateWithAggregate` by appending a literal `1` to the grouping key list of the resulting `Aggregate` operator when the input plan contains zero output columns. In this way, `spark.emptyDataFrame.dropDuplicates()` is now translated into a grouping aggregation, roughly depicted as:
```scala
spark.emptyDataFrame.dropDuplicates()
=> spark.emptyDataFrame.groupBy(lit(1)).agg(Map.empty[String, String])
```
Which is now properly treated as a grouping aggregation and returns the correct answer.
## How was this patch tested?
New unit tests added
Author: Feng Liu <fengliu@databricks.com>
Closes#20174 from liufengdb/fix-duplicate.
## What changes were proposed in this pull request?
This is mostly from https://github.com/apache/spark/pull/13775
The wrapper solution is pretty good for string/binary type, as the ORC column vector doesn't keep bytes in a continuous memory region, and has a significant overhead when copying the data to Spark columnar batch. For other cases, the wrapper solution is almost same with the current solution.
I think we can treat the wrapper solution as a baseline and keep improving the writing to Spark solution.
## How was this patch tested?
existing tests.
Author: Wenchen Fan <wenchen@databricks.com>
Closes#20205 from cloud-fan/orc.
## What changes were proposed in this pull request?
This pr fixed the issue when casting structs into strings;
```
scala> val df = Seq(((1, "a"), 0), ((2, "b"), 0)).toDF("a", "b")
scala> df.write.saveAsTable("t")
scala> sql("SELECT CAST(a AS STRING) FROM t").show
+-------------------+
| a|
+-------------------+
|[0,1,1800000001,61]|
|[0,2,1800000001,62]|
+-------------------+
```
This pr modified the result into;
```
+------+
| a|
+------+
|[1, a]|
|[2, b]|
+------+
```
## How was this patch tested?
Added tests in `CastSuite`.
Author: Takeshi Yamamuro <yamamuro@apache.org>
Closes#20176 from maropu/SPARK-22981.
## What changes were proposed in this pull request?
This PR adds an ORC columnar-batch reader to native `OrcFileFormat`. Since both Spark `ColumnarBatch` and ORC `RowBatch` are used together, it is faster than the current Spark implementation. This replaces the prior PR, #17924.
Also, this PR adds `OrcReadBenchmark` to show the performance improvement.
## How was this patch tested?
Pass the existing test cases.
Author: Dongjoon Hyun <dongjoon@apache.org>
Closes#19943 from dongjoon-hyun/SPARK-16060.
## What changes were proposed in this pull request?
This PR fixes a longstanding correctness bug in `GenerateUnsafeRowJoiner`. This class was introduced in https://github.com/apache/spark/pull/7821 (July 2015 / Spark 1.5.0+) and is used to combine pairs of UnsafeRows in TungstenAggregationIterator, CartesianProductExec, and AppendColumns.
### Bugs fixed by this patch
1. **Incorrect combining of null-tracking bitmaps**: when concatenating two UnsafeRows, the implementation "Concatenate the two bitsets together into a single one, taking padding into account". If one row has no columns then it has a bitset size of 0, but the code was incorrectly assuming that if the left row had a non-zero number of fields then the right row would also have at least one field, so it was copying invalid bytes and and treating them as part of the bitset. I'm not sure whether this bug was also present in the original implementation or whether it was introduced in https://github.com/apache/spark/pull/7892 (which fixed another bug in this code).
2. **Incorrect updating of data offsets for null variable-length fields**: after updating the bitsets and copying fixed-length and variable-length data, we need to perform adjustments to the offsets pointing the start of variable length fields's data. The existing code was _conditionally_ adding a fixed offset to correct for the new length of the combined row, but it is unsafe to do this if the variable-length field has a null value: we always represent nulls by storing `0` in the fixed-length slot, but this code was incorrectly incrementing those values. This bug was present since the original version of `GenerateUnsafeRowJoiner`.
### Why this bug remained latent for so long
The PR which introduced `GenerateUnsafeRowJoiner` features several randomized tests, including tests of the cases where one side of the join has no fields and where string-valued fields are null. However, the existing assertions were too weak to uncover this bug:
- If a null field has a non-zero value in its fixed-length data slot then this will not cause problems for field accesses because the null-tracking bitmap should still be correct and we will not try to use the incorrect offset for anything.
- If the null tracking bitmap is corrupted by joining against a row with no fields then the corruption occurs in field numbers past the actual field numbers contained in the row. Thus valid `isNullAt()` calls will not read the incorrectly-set bits.
The existing `GenerateUnsafeRowJoinerSuite` tests only exercised `.get()` and `isNullAt()`, but didn't actually check the UnsafeRows for bit-for-bit equality, preventing these bugs from failing assertions. It turns out that there was even a [GenerateUnsafeRowJoinerBitsetSuite](03377d2522/sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/expressions/codegen/GenerateUnsafeRowJoinerBitsetSuite.scala) but it looks like it also didn't catch this problem because it only tested the bitsets in an end-to-end fashion by accessing them through the `UnsafeRow` interface instead of actually comparing the bitsets' bytes.
### Impact of these bugs
- This bug will cause `equals()` and `hashCode()` to be incorrect for these rows, which will be problematic in case`GenerateUnsafeRowJoiner`'s results are used as join or grouping keys.
- Chained / repeated invocations of `GenerateUnsafeRowJoiner` may result in reads from invalid null bitmap positions causing fields to incorrectly become NULL (see the end-to-end example below).
- It looks like this generally only happens in `CartesianProductExec`, which our query optimizer often avoids executing (usually we try to plan a `BroadcastNestedLoopJoin` instead).
### End-to-end test case demonstrating the problem
The following query demonstrates how this bug may result in incorrect query results:
```sql
set spark.sql.autoBroadcastJoinThreshold=-1; -- Needed to trigger CartesianProductExec
create table a as select * from values 1;
create table b as select * from values 2;
SELECT
t3.col1,
t1.col1
FROM a t1
CROSS JOIN b t2
CROSS JOIN b t3
```
This should return `(2, 1)` but instead was returning `(null, 1)`.
Column pruning ends up trimming off all columns from `t2`, so when `t2` joins with another table this triggers the bitmap-copying bug. This incorrect bitmap is subsequently copied again when performing the final join, causing the final output to have an incorrectly-set null bit for the first field.
## How was this patch tested?
Strengthened the assertions in existing tests in GenerateUnsafeRowJoinerSuite. Also verified that the end-to-end test case which uncovered this now passes.
Author: Josh Rosen <joshrosen@databricks.com>
Closes#20181 from JoshRosen/SPARK-22984-fix-generate-unsaferow-joiner-bitmap-bugs.
## What changes were proposed in this pull request?
**The current shuffle planning logic**
1. Each operator specifies the distribution requirements for its children, via the `Distribution` interface.
2. Each operator specifies its output partitioning, via the `Partitioning` interface.
3. `Partitioning.satisfy` determines whether a `Partitioning` can satisfy a `Distribution`.
4. For each operator, check each child of it, add a shuffle node above the child if the child partitioning can not satisfy the required distribution.
5. For each operator, check if its children's output partitionings are compatible with each other, via the `Partitioning.compatibleWith`.
6. If the check in 5 failed, add a shuffle above each child.
7. try to eliminate the shuffles added in 6, via `Partitioning.guarantees`.
This design has a major problem with the definition of "compatible".
`Partitioning.compatibleWith` is not well defined, ideally a `Partitioning` can't know if it's compatible with other `Partitioning`, without more information from the operator. For example, `t1 join t2 on t1.a = t2.b`, `HashPartitioning(a, 10)` should be compatible with `HashPartitioning(b, 10)` under this case, but the partitioning itself doesn't know it.
As a result, currently `Partitioning.compatibleWith` always return false except for literals, which make it almost useless. This also means, if an operator has distribution requirements for multiple children, Spark always add shuffle nodes to all the children(although some of them can be eliminated). However, there is no guarantee that the children's output partitionings are compatible with each other after adding these shuffles, we just assume that the operator will only specify `ClusteredDistribution` for multiple children.
I think it's very hard to guarantee children co-partition for all kinds of operators, and we can not even give a clear definition about co-partition between distributions like `ClusteredDistribution(a,b)` and `ClusteredDistribution(c)`.
I think we should drop the "compatible" concept in the distribution model, and let the operator achieve the co-partition requirement by special distribution requirements.
**Proposed shuffle planning logic after this PR**
(The first 4 are same as before)
1. Each operator specifies the distribution requirements for its children, via the `Distribution` interface.
2. Each operator specifies its output partitioning, via the `Partitioning` interface.
3. `Partitioning.satisfy` determines whether a `Partitioning` can satisfy a `Distribution`.
4. For each operator, check each child of it, add a shuffle node above the child if the child partitioning can not satisfy the required distribution.
5. For each operator, check if its children's output partitionings have the same number of partitions.
6. If the check in 5 failed, pick the max number of partitions from children's output partitionings, and add shuffle to child whose number of partitions doesn't equal to the max one.
The new distribution model is very simple, we only have one kind of relationship, which is `Partitioning.satisfy`. For multiple children, Spark only guarantees they have the same number of partitions, and it's the operator's responsibility to leverage this guarantee to achieve more complicated requirements. For example, non-broadcast joins can use the newly added `HashPartitionedDistribution` to achieve co-partition.
## How was this patch tested?
existing tests.
Author: Wenchen Fan <wenchen@databricks.com>
Closes#19080 from cloud-fan/exchange.
## What changes were proposed in this pull request?
The following SQL query should return zero rows, but in Spark it actually returns one row:
```
SELECT 1 from (
SELECT 1 AS z,
MIN(a.x)
FROM (select 1 as x) a
WHERE false
) b
where b.z != b.z
```
The problem stems from the `PushDownPredicate` rule: when this rule encounters a filter on top of an Aggregate operator, e.g. `Filter(Agg(...))`, it removes the original filter and adds a new filter onto Aggregate's child, e.g. `Agg(Filter(...))`. This is sometimes okay, but the case above is a counterexample: because there is no explicit `GROUP BY`, we are implicitly computing a global aggregate over the entire table so the original filter was not acting like a `HAVING` clause filtering the number of groups: if we push this filter then it fails to actually reduce the cardinality of the Aggregate output, leading to the wrong answer.
In 2016 I fixed a similar problem involving invalid pushdowns of data-independent filters (filters which reference no columns of the filtered relation). There was additional discussion after my fix was merged which pointed out that my patch was an incomplete fix (see #15289), but it looks I must have either misunderstood the comment or forgot to follow up on the additional points raised there.
This patch fixes the problem by choosing to never push down filters in cases where there are no grouping expressions. Since there are no grouping keys, the only columns are aggregate columns and we can't push filters defined over aggregate results, so this change won't cause us to miss out on any legitimate pushdown opportunities.
## How was this patch tested?
New regression tests in `SQLQueryTestSuite` and `FilterPushdownSuite`.
Author: Josh Rosen <joshrosen@databricks.com>
Closes#20180 from JoshRosen/SPARK-22983-dont-push-filters-beneath-aggs-with-empty-grouping-expressions.
