## 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?
Currently `PartitioningAwareFileIndex` accepts an optional parameter `userPartitionSchema`. If provided, it will combine the inferred partition schema with the parameter.
However,
1. to get `userPartitionSchema`, we need to combine inferred partition schema with `userSpecifiedSchema`
2. to get the inferred partition schema, we have to create a temporary file index.
Only after that, a final version of `PartitioningAwareFileIndex` can be created.
This can be improved by passing `userSpecifiedSchema` to `PartitioningAwareFileIndex`.
With the improvement, we can reduce redundant code and avoid parsing the file partition twice.
## How was this patch tested?
Unit test
Author: Gengliang Wang <gengliang.wang@databricks.com>
Closes#21004 from gengliangwang/PartitioningAwareFileIndex.
## 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?
Many suites currently leak Spark sessions (sometimes with stopped SparkContexts) via the thread-local active Spark session and default Spark session. We should attempt to clean these up and detect when this happens to improve the reproducibility of tests.
## How was this patch tested?
Existing tests
Author: Eric Liang <ekl@databricks.com>
Closes#21058 from ericl/clear-session.
## What changes were proposed in this pull request?
This PR proposes to add `collect` to a query executor as an action.
Seems `collect` / `collect` with Arrow are not recognised via `QueryExecutionListener` as an action. For example, if we have a custom listener as below:
```scala
package org.apache.spark.sql
import org.apache.spark.internal.Logging
import org.apache.spark.sql.execution.QueryExecution
import org.apache.spark.sql.util.QueryExecutionListener
class TestQueryExecutionListener extends QueryExecutionListener with Logging {
override def onSuccess(funcName: String, qe: QueryExecution, durationNs: Long): Unit = {
logError("Look at me! I'm 'onSuccess'")
}
override def onFailure(funcName: String, qe: QueryExecution, exception: Exception): Unit = { }
}
```
and set `spark.sql.queryExecutionListeners` to `org.apache.spark.sql.TestQueryExecutionListener`
Other operations in PySpark or Scala side seems fine:
```python
>>> sql("SELECT * FROM range(1)").show()
```
```
18/04/09 17:02:04 ERROR TestQueryExecutionListener: Look at me! I'm 'onSuccess'
+---+
| id|
+---+
| 0|
+---+
```
```scala
scala> sql("SELECT * FROM range(1)").collect()
```
```
18/04/09 16:58:41 ERROR TestQueryExecutionListener: Look at me! I'm 'onSuccess'
res1: Array[org.apache.spark.sql.Row] = Array([0])
```
but ..
**Before**
```python
>>> sql("SELECT * FROM range(1)").collect()
```
```
[Row(id=0)]
```
```python
>>> spark.conf.set("spark.sql.execution.arrow.enabled", "true")
>>> sql("SELECT * FROM range(1)").toPandas()
```
```
id
0 0
```
**After**
```python
>>> sql("SELECT * FROM range(1)").collect()
```
```
18/04/09 16:57:58 ERROR TestQueryExecutionListener: Look at me! I'm 'onSuccess'
[Row(id=0)]
```
```python
>>> spark.conf.set("spark.sql.execution.arrow.enabled", "true")
>>> sql("SELECT * FROM range(1)").toPandas()
```
```
18/04/09 17:53:26 ERROR TestQueryExecutionListener: Look at me! I'm 'onSuccess'
id
0 0
```
## How was this patch tested?
I have manually tested as described above and unit test was added.
Author: hyukjinkwon <gurwls223@apache.org>
Closes#21007 from HyukjinKwon/SPARK-23942.
## 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.
SQLMetricsTestUtils.currentExecutionIds() was racing with the listener
bus, which lead to some flaky tests. We should wait till the listener bus is
empty.
I tested by adding some Thread.sleep()s in SQLAppStatusListener, which
reproduced the exceptions I saw on Jenkins. With this change, they went
away.
Author: Imran Rashid <irashid@cloudera.com>
Closes#21041 from squito/SPARK-23962.
## What changes were proposed in this pull request?
Mark `HashAggregateExec.bufVars` as transient to avoid it from being serialized.
Also manually null out this field at the end of `doProduceWithoutKeys()` to shorten its lifecycle, because it'll no longer be used after that.
## How was this patch tested?
Existing tests.
Author: Kris Mok <kris.mok@databricks.com>
Closes#21039 from rednaxelafx/codegen-improve.
## 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?
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?
Proposed tests checks that only subset of input dataset is touched during schema inferring.
Author: Maxim Gekk <maxim.gekk@databricks.com>
Closes#20963 from MaxGekk/json-sampling-tests.
## What changes were proposed in this pull request?
Column.scala and Functions.scala have asc_nulls_first, asc_nulls_last, desc_nulls_first and desc_nulls_last. Add the corresponding python APIs in column.py and functions.py
## How was this patch tested?
Add doctest
Author: Huaxin Gao <huaxing@us.ibm.com>
Closes#20962 from huaxingao/spark-23847.
## What changes were proposed in this pull request?
Add docstring to clarify default window frame boundaries with and without orderBy clause
## How was this patch tested?
Manually generate doc and check.
Author: Li Jin <ice.xelloss@gmail.com>
Closes#20978 from icexelloss/SPARK-23861-window-doc.
## What changes were proposed in this pull request?
This pull request tries to improve the error message for spark while reading parquet files with different schemas, e.g. One with a STRING column and the other with a INT column. A new ParquetSchemaColumnConvertNotSupportedException is added to replace the old UnsupportedOperationException. The Exception is again wrapped in FileScanRdd.scala to throw a more a general QueryExecutionException with the actual parquet file name which trigger the exception.
## How was this patch tested?
Unit tests added to check the new exception and verify the error messages.
Also manually tested with two parquet with different schema to check the error message.
<img width="1125" alt="screen shot 2018-03-30 at 4 03 04 pm" src="https://user-images.githubusercontent.com/37087310/38156580-dd58a140-3433-11e8-973a-b816d859fbe1.png">
Author: Yuchen Huo <yuchen.huo@databricks.com>
Closes#20953 from yuchenhuo/SPARK-23822.
## 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.
## 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?
Currently, the active spark session is set inconsistently (e.g., in createDataFrame, prior to query execution). Many places in spark also incorrectly query active session when they should be calling activeSession.getOrElse(defaultSession) and so might get None even if a Spark session exists.
The semantics here can be cleaned up if we also set the active session when the default session is set.
Related: https://github.com/apache/spark/pull/20926/files
## How was this patch tested?
Unit test, existing test. Note that if https://github.com/apache/spark/pull/20926 merges first we should also update the tests there.
Author: Eric Liang <ekl@databricks.com>
Closes#20927 from ericl/active-session-cleanup.
## What changes were proposed in this pull request?
Migrate foreach sink to DataSourceV2.
Since the previous attempt at this PR #20552, we've changed and strictly defined the lifecycle of writer components. This means we no longer need the complicated lifecycle shim from that PR; it just naturally works.
## How was this patch tested?
existing tests
Author: Jose Torres <torres.joseph.f+github@gmail.com>
Closes#20951 from jose-torres/foreach.
## 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?
Currently, the requiredChildDistribution does not specify the partitions. This can cause the weird corner cases where the child's distribution is `SinglePartition` which satisfies the required distribution of `ClusterDistribution(no-num-partition-requirement)`, thus eliminating the shuffle needed to repartition input data into the required number of partitions (i.e. same as state stores). That can lead to "file not found" errors on the state store delta files as the micro-batch-with-no-shuffle will not run certain tasks and therefore not generate the expected state store delta files.
This PR adds the required constraint on the number of partitions.
## How was this patch tested?
Modified test harness to always check that ANY stateful operator should have a constraint on the number of partitions. As part of that, the existing opt-in checks on child output partitioning were removed, as they are redundant.
Author: Tathagata Das <tathagata.das1565@gmail.com>
Closes#20941 from tdas/SPARK-23827.
## 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?
Roll forward c68ec4e (#20688).
There are two minor test changes required:
* An error which used to be TreeNodeException[ArithmeticException] is no longer wrapped and is now just ArithmeticException.
* The test framework simply does not set the active Spark session. (Or rather, it doesn't do so early enough - I think it only happens when a query is analyzed.) I've added the required logic to SQLTestUtils.
## How was this patch tested?
existing tests
Author: Jose Torres <torres.joseph.f+github@gmail.com>
Author: jerryshao <sshao@hortonworks.com>
Closes#20922 from jose-torres/ratefix.
## 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?
Set default Spark session in the TestSparkSession and TestHiveSparkSession constructors.
## How was this patch tested?
new unit tests
Author: Jose Torres <torres.joseph.f+github@gmail.com>
Closes#20926 from jose-torres/test3.
## 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?
This PR migrate micro batch rate source to V2 API and rewrite UTs to suite V2 test.
## How was this patch tested?
UTs.
Author: jerryshao <sshao@hortonworks.com>
Closes#20688 from jerryshao/SPARK-23096.
## 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?
This pr added TPCDS v2.7 (latest) queries in `TPCDSQuerySuite` because the current `TPCDSQuerySuite` tests older one (v1.4) and some queries are different from v1.4 and v2.7. Since the original v2.7 queries have the syntaxes that Spark cannot parse, I changed these queries in a following way:
- [date] + 14 days -> date + `INTERVAL` 14 days
- [column name] as "30 days" -> [column name] as \`30 days\`
- Fix some syntax errors, e.g., missing brackets
## How was this patch tested?
Added tests in `TPCDSQuerySuite`.
Author: Takeshi Yamamuro <yamamuro@apache.org>
Closes#20343 from maropu/TPCDSV2_7.
## What changes were proposed in this pull request?
The serializability test uses the same MemoryStream instance for 3 different queries. If any of those queries ask it to commit before the others have run, the rest will see empty dataframes. This can fail the test if q3 is affected.
We should use one instance per query instead.
## How was this patch tested?
Existing unit test. If I move q2.processAllAvailable() before starting q3, the test always fails without the fix.
Author: Jose Torres <torres.joseph.f+github@gmail.com>
Closes#20896 from jose-torres/fixrace.
## What changes were proposed in this pull request?
We should provide customized canonicalize plan for `InMemoryRelation` and `InMemoryTableScanExec`. Otherwise, we can wrongly treat two different cached plans as same result. It causes wrongly reused exchange then.
For a test query like this:
```scala
val cached = spark.createDataset(Seq(TestDataUnion(1, 2, 3), TestDataUnion(4, 5, 6))).cache()
val group1 = cached.groupBy("x").agg(min(col("y")) as "value")
val group2 = cached.groupBy("x").agg(min(col("z")) as "value")
group1.union(group2)
```
Canonicalized plans before:
First exchange:
```
Exchange hashpartitioning(none#0, 5)
+- *(1) HashAggregate(keys=[none#0], functions=[partial_min(none#1)], output=[none#0, none#4])
+- *(1) InMemoryTableScan [none#0, none#1]
+- InMemoryRelation [x#4253, y#4254, z#4255], true, 10000, StorageLevel(disk, memory, deserialized, 1 replicas)
+- LocalTableScan [x#4253, y#4254, z#4255]
```
Second exchange:
```
Exchange hashpartitioning(none#0, 5)
+- *(3) HashAggregate(keys=[none#0], functions=[partial_min(none#1)], output=[none#0, none#4])
+- *(3) InMemoryTableScan [none#0, none#1]
+- InMemoryRelation [x#4253, y#4254, z#4255], true, 10000, StorageLevel(disk, memory, deserialized, 1 replicas)
+- LocalTableScan [x#4253, y#4254, z#4255]
```
You can find that they have the canonicalized plans are the same, although we use different columns in two `InMemoryTableScan`s.
Canonicalized plan after:
First exchange:
```
Exchange hashpartitioning(none#0, 5)
+- *(1) HashAggregate(keys=[none#0], functions=[partial_min(none#1)], output=[none#0, none#4])
+- *(1) InMemoryTableScan [none#0, none#1]
+- InMemoryRelation [none#0, none#1, none#2], true, 10000, StorageLevel(memory, 1 replicas)
+- LocalTableScan [none#0, none#1, none#2]
```
Second exchange:
```
Exchange hashpartitioning(none#0, 5)
+- *(3) HashAggregate(keys=[none#0], functions=[partial_min(none#1)], output=[none#0, none#4])
+- *(3) InMemoryTableScan [none#0, none#2]
+- InMemoryRelation [none#0, none#1, none#2], true, 10000, StorageLevel(memory, 1 replicas)
+- LocalTableScan [none#0, none#1, none#2]
```
## How was this patch tested?
Added unit test.
Author: Liang-Chi Hsieh <viirya@gmail.com>
Closes#20831 from viirya/SPARK-23614.
## 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?
Currently we allow writing data frames with empty schema into a file based datasource for certain file formats such as JSON, ORC etc. For formats such as Parquet and Text, we raise error at different times of execution. For text format, we return error from the driver early on in processing where as for format such as parquet, the error is raised from executor.
**Example**
spark.emptyDataFrame.write.format("parquet").mode("overwrite").save(path)
**Results in**
``` SQL
org.apache.parquet.schema.InvalidSchemaException: Cannot write a schema with an empty group: message spark_schema {
}
at org.apache.parquet.schema.TypeUtil$1.visit(TypeUtil.java:27)
at org.apache.parquet.schema.TypeUtil$1.visit(TypeUtil.java:37)
at org.apache.parquet.schema.MessageType.accept(MessageType.java:58)
at org.apache.parquet.schema.TypeUtil.checkValidWriteSchema(TypeUtil.java:23)
at org.apache.parquet.hadoop.ParquetFileWriter.<init>(ParquetFileWriter.java:225)
at org.apache.parquet.hadoop.ParquetOutputFormat.getRecordWriter(ParquetOutputFormat.java:342)
at org.apache.parquet.hadoop.ParquetOutputFormat.getRecordWriter(ParquetOutputFormat.java:302)
at org.apache.spark.sql.execution.datasources.parquet.ParquetOutputWriter.<init>(ParquetOutputWriter.scala:37)
at org.apache.spark.sql.execution.datasources.parquet.ParquetFileFormat$$anon$1.newInstance(ParquetFileFormat.scala:151)
at org.apache.spark.sql.execution.datasources.FileFormatWriter$SingleDirectoryWriteTask.newOutputWriter(FileFormatWriter.scala:376)
at org.apache.spark.sql.execution.datasources.FileFormatWriter$SingleDirectoryWriteTask.execute(FileFormatWriter.scala:387)
at org.apache.spark.sql.execution.datasources.FileFormatWriter$$anonfun$org$apache$spark$sql$execution$datasources$FileFormatWriter$$executeTask$3.apply(FileFormatWriter.scala:278)
at org.apache.spark.sql.execution.datasources.FileFormatWriter$$anonfun$org$apache$spark$sql$execution$datasources$FileFormatWriter$$executeTask$3.apply(FileFormatWriter.scala:276)
at org.apache.spark.util.Utils$.tryWithSafeFinallyAndFailureCallbacks(Utils.scala:1411)
at org.apache.spark.sql.execution.datasources.FileFormatWriter$.org$apache$spark$sql$execution$datasources$FileFormatWriter$$executeTask(FileFormatWriter.scala:281)
at org.apache.spark.sql.execution.datasources.FileFormatWriter$$anonfun$write$1.apply(FileFormatWriter.scala:206)
at org.apache.spark.sql.execution.datasources.FileFormatWriter$$anonfun$write$1.apply(FileFormatWriter.scala:205)
at org.apache.spark.scheduler.ResultTask.runTask(ResultTask.scala:87)
at org.apache.spark.scheduler.Task.run(Task.scala:109)
at org.apache.spark.executor.Executor$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.
```
In this PR, we unify the error processing and raise error on attempt to write empty schema based dataframes into file based datasource (orc, parquet, text , csv, json etc) early on in the processing.
## How was this patch tested?
Unit tests added in FileBasedDatasourceSuite.
Author: Dilip Biswal <dbiswal@us.ibm.com>
Closes#20579 from dilipbiswal/spark-23372.
## What changes were proposed in this pull request?
Output metrics were not filled when parquet sink used.
This PR fixes this problem by passing a `BasicWriteJobStatsTracker` in `FileStreamSink`.
## How was this patch tested?
Additional unit test added.
Author: Gabor Somogyi <gabor.g.somogyi@gmail.com>
Closes#20745 from gaborgsomogyi/SPARK-23288.
## What changes were proposed in this pull request?
To fix `scala.MatchError` in `literals.sql.out`, this pr added an entry for `CalendarIntervalType` in `QueryExecution.toHiveStructString`.
## How was this patch tested?
Existing tests and added tests in `literals.sql`
Author: Takeshi Yamamuro <yamamuro@apache.org>
Closes#20872 from maropu/FixIntervalTests.
## 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.
## How was this patch tested?
Manual tests and unit tests were added.
Author: hyukjinkwon <gurwls223@gmail.com>
Closes#20727 from HyukjinKwon/linesep-text.
## 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?
Report SinglePartition in DataSourceV2ScanExec when there's exactly 1 data reader factory.
Note that this means reader factories end up being constructed as partitioning is checked; let me know if you think that could be a problem.
## How was this patch tested?
existing unit tests
Author: Jose Torres <jose@databricks.com>
Author: Jose Torres <torres.joseph.f+github@gmail.com>
Closes#20726 from jose-torres/SPARK-23574.
## What changes were proposed in this pull request?
Currently, some tests have an assumption that `spark.sql.sources.default=parquet`. In fact, that is a correct assumption, but that assumption makes it difficult to test new data source format.
This PR aims to
- Improve test suites more robust and makes it easy to test new data sources in the future.
- Test new native ORC data source with the full existing Apache Spark test coverage.
As an example, the PR uses `spark.sql.sources.default=orc` during reviews. The value should be `parquet` when this PR is accepted.
## How was this patch tested?
Pass the Jenkins with updated tests.
Author: Dongjoon Hyun <dongjoon@apache.org>
Closes#20705 from dongjoon-hyun/SPARK-23553.
Clean up SparkPlanGraphWrapper objects from InMemoryStore together with cleaning up SQLExecutionUIData
existing unit test was extended to check also SparkPlanGraphWrapper object count
vanzin
Author: myroslavlisniak <acnipin@gmail.com>
Closes#20813 from myroslavlisniak/master.
## What changes were proposed in this pull request?
As discussion in #20675, we need add a new interface `ContinuousDataReaderFactory` to support the requirements of setting start offset in Continuous Processing.
## How was this patch tested?
Existing UT.
Author: Yuanjian Li <xyliyuanjian@gmail.com>
Closes#20689 from xuanyuanking/SPARK-23533.
## What changes were proposed in this pull request?
Revise doc of method pushFilters in SupportsPushDownFilters/SupportsPushDownCatalystFilters
In `FileSourceStrategy`, except `partitionKeyFilters`(the references of which is subset of partition keys), all filters needs to be evaluated after scanning. Otherwise, Spark will get wrong result from data sources like Orc/Parquet.
This PR is to improve the doc.
Author: Wang Gengliang <gengliang.wang@databricks.com>
Closes#20769 from gengliangwang/revise_pushdown_doc.