lanyuanxiaoyao/hudi
1
0
Fork 0
Code Activity

[HUDI-4367] Support copyToTable on call (#6054)

Browse Source
This commit is contained in:
苏承祥
2022-07-08 15:08:11 +08:00
committed by GitHub
parent a998586396
commit f20acb8dc3
3 changed files with 482 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

128
hudi-spark-datasource/hudi-spark/src/main/scala/org/apache/spark/sql/hudi/command/procedures/CopyToTableProcedure.scala Normal file
View File

@@ -0,0 +1,128 @@
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You under the Apache License, Version 2.0
* (the "License"); you may not use this file except in compliance with
* the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.apache.spark.sql.hudi.command.procedures
import org.apache.hudi.DataSourceReadOptions
import org.apache.spark.internal.Logging
import org.apache.spark.sql.types.{DataTypes, Metadata, StructField, StructType}
import org.apache.spark.sql.{Row, SaveMode}
import java.util.function.Supplier
class CopyToTableProcedure extends BaseProcedure with ProcedureBuilder with Logging {
private val PARAMETERS = Array[ProcedureParameter](
ProcedureParameter.required(0, "table", DataTypes.StringType, None),
ProcedureParameter.optional(1, "query_type", DataTypes.StringType, DataSourceReadOptions.QUERY_TYPE_SNAPSHOT_OPT_VAL),
ProcedureParameter.required(2, "new_table", DataTypes.StringType, None),
ProcedureParameter.optional(3, "begin_instance_time", DataTypes.StringType, ""),
ProcedureParameter.optional(4, "end_instance_time", DataTypes.StringType, ""),
ProcedureParameter.optional(5, "as_of_instant", DataTypes.StringType, ""),
ProcedureParameter.optional(6, "save_mode", DataTypes.StringType, "overwrite")
)
private val OUTPUT_TYPE = new StructType(Array[StructField](
StructField("status", DataTypes.IntegerType, nullable = true, Metadata.empty))
)
def parameters: Array[ProcedureParameter] = PARAMETERS
def outputType: StructType = OUTPUT_TYPE
override def call(args: ProcedureArgs): Seq[Row] = {
super.checkArgs(PARAMETERS, args)
val tableName = getArgValueOrDefault(args, PARAMETERS(0))
val queryType = getArgValueOrDefault(args, PARAMETERS(1)).get.asInstanceOf[String]
val newTableName = getArgValueOrDefault(args, PARAMETERS(2)).get.asInstanceOf[String]
val beginInstance = getArgValueOrDefault(args, PARAMETERS(3)).get.asInstanceOf[String]
val endInstance = getArgValueOrDefault(args, PARAMETERS(4)).get.asInstanceOf[String]
val asOfInstant = getArgValueOrDefault(args, PARAMETERS(5)).get.asInstanceOf[String]
val saveModeStr = getArgValueOrDefault(args, PARAMETERS(6)).get.asInstanceOf[String]
assert(saveModeStr.nonEmpty, "save_mode(append,overwrite) can not be null.")
val saveMode: Any = saveModeStr.toLowerCase match {
case "overwrite" => SaveMode.Overwrite
case "append" => SaveMode.Append
case _ => assert(assertion = false, s"save_mode not support $saveModeStr.")
}
val tablePath = getBasePath(tableName)
val sourceDataFrame = queryType match {
case DataSourceReadOptions.QUERY_TYPE_SNAPSHOT_OPT_VAL => if (asOfInstant.nonEmpty) {
sparkSession.read
.format("org.apache.hudi")
.option(DataSourceReadOptions.QUERY_TYPE.key, DataSourceReadOptions.QUERY_TYPE_SNAPSHOT_OPT_VAL)
.option(DataSourceReadOptions.TIME_TRAVEL_AS_OF_INSTANT.key, asOfInstant)
.load(tablePath)
} else {
sparkSession.read
.format("org.apache.hudi")
.option(DataSourceReadOptions.QUERY_TYPE.key, DataSourceReadOptions.QUERY_TYPE_SNAPSHOT_OPT_VAL)
.load(tablePath)
}
case DataSourceReadOptions.QUERY_TYPE_INCREMENTAL_OPT_VAL =>
assert(beginInstance.nonEmpty && endInstance.nonEmpty, "when the query_type is incremental, begin_instance_time and end_instance_time can not be null.")
sparkSession.read
.format("org.apache.hudi")
.option(DataSourceReadOptions.QUERY_TYPE.key, DataSourceReadOptions.QUERY_TYPE_INCREMENTAL_OPT_VAL)
.option(DataSourceReadOptions.BEGIN_INSTANTTIME.key, beginInstance)
.option(DataSourceReadOptions.END_INSTANTTIME.key, endInstance)
.load(tablePath)
case DataSourceReadOptions.QUERY_TYPE_READ_OPTIMIZED_OPT_VAL =>
sparkSession.read
.format("org.apache.hudi")
.option(DataSourceReadOptions.QUERY_TYPE.key, DataSourceReadOptions.QUERY_TYPE_READ_OPTIMIZED_OPT_VAL)
.load(tablePath)
}
if (sparkSession.catalog.tableExists(newTableName)) {
val schema = sparkSession.read.table(newTableName).schema
val selectColumns = schema.fields.toStream.map(_.name)
sourceDataFrame.selectExpr(selectColumns: _*)
.write
.mode(saveMode.toString)
.saveAsTable(newTableName)
} else {
sourceDataFrame.write
.mode(saveMode.toString)
.saveAsTable(newTableName)
}
Seq(Row(0))
}
override def build = new CopyToTableProcedure()
}
object CopyToTableProcedure {
val NAME = "copy_to_table"
def builder: Supplier[ProcedureBuilder] = new Supplier[ProcedureBuilder] {
override def get() = new CopyToTableProcedure()
}
}

1
hudi-spark-datasource/hudi-spark/src/main/scala/org/apache/spark/sql/hudi/command/procedures/HoodieProcedures.scala
View File

@@ -74,6 +74,7 @@ object HoodieProcedures {
mapBuilder.put(ShowMetadataStatsProcedure.NAME, ShowMetadataStatsProcedure.builder)
mapBuilder.put(ValidateMetadataFilesProcedure.NAME, ValidateMetadataFilesProcedure.builder)
mapBuilder.put(ShowFsPathDetailProcedure.NAME, ShowFsPathDetailProcedure.builder)
mapBuilder.put(CopyToTableProcedure.NAME, CopyToTableProcedure.builder)
mapBuilder.build
}
}

353
hudi-spark-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/procedure/TestCopyToTableProcedure.scala Normal file
View File

@@ -0,0 +1,353 @@
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You under the Apache License, Version 2.0
* (the "License"); you may not use this file except in compliance with
* the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.apache.spark.sql.hudi.procedure
import org.apache.spark.sql.Row
import org.apache.spark.sql.hudi.HoodieSparkSqlTestBase
import java.util
class TestCopyToTableProcedure extends HoodieSparkSqlTestBase {
test("Test Call copy_to_table Procedure with default params") {
withTempDir { tmp =>
val tableName = generateTableName
// create table
spark.sql(
s"""
|create table $tableName (
| id int,
| name string,
| price double,
| ts long
|) using hudi
| location '${tmp.getCanonicalPath}/$tableName'
| tblproperties (
| primaryKey = 'id',
| preCombineField = 'ts'
| )
""".stripMargin)
// insert data to table
spark.sql(s"insert into $tableName select 1, 'a1', 10, 1000")
spark.sql(s"insert into $tableName select 2, 'a2', 20, 1500")
spark.sql(s"insert into $tableName select 3, 'a3', 30, 2000")
spark.sql(s"insert into $tableName select 4, 'a4', 40, 2500")
val copyTableName = generateTableName
// Check required fields
checkExceptionContain(s"call copy_to_table(table=>'$tableName')")(s"Argument: new_table is required")
val row = spark.sql(s"""call copy_to_table(table=>'$tableName',new_table=>'$copyTableName')""").collectAsList()
assert(row.size() == 1 && row.get(0).get(0) == 0)
val copyTableCount = spark.sql(s"""select count(1) from $copyTableName""").collectAsList()
assert(copyTableCount.size() == 1 && copyTableCount.get(0).get(0) == 4)
}
}
test("Test Call copy_to_table Procedure with snapshot") {
withTempDir { tmp =>
val tableName = generateTableName
// create table
spark.sql(
s"""
|create table $tableName (
| id int,
| name string,
| price double,
| ts long
|) using hudi
| location '${tmp.getCanonicalPath}/$tableName'
| tblproperties (
| primaryKey = 'id',
| preCombineField = 'ts'
| )
""".stripMargin)
// insert 4 rows data to table
spark.sql(s"insert into $tableName select 1, 'a1', 10, 1000")
spark.sql(s"insert into $tableName select 2, 'a2', 20, 1500")
spark.sql(s"insert into $tableName select 3, 'a3', 30, 2000")
spark.sql(s"insert into $tableName select 4, 'a4', 40, 2500")
val copyTableName = generateTableName
val row = spark.sql(s"""call copy_to_table(table=>'$tableName',new_table=>'$copyTableName',query_type=>'snapshot')""").collectAsList()
// check exit code
assert(row.size() == 1 && row.get(0).get(0) == 0)
val copyTableCount = spark.sql(s"""select count(1) from $copyTableName""").collectAsList()
assert(copyTableCount.size() == 1 && copyTableCount.get(0).get(0) == 4)
// insert 4 rows data to table
spark.sql(s"insert into $tableName select 5, 'a1', 10, 1000")
// mark max instanceTime.total row is 5
val instanceTime = spark.sql(s"select max(_hoodie_commit_time) from $tableName").collectAsList().get(0).get(0)
spark.sql(s"insert into $tableName select 6, 'a1', 10, 1000")
spark.sql(s"insert into $tableName select 7, 'a1', 10, 1000")
spark.sql(s"insert into $tableName select 8, 'a1', 10, 1000")
// check row count after twice insert
val finalTableCount = spark.sql(s"select * from $tableName").count()
assert(finalTableCount == 8)
// check snapshot copy with mark instanceTime
val copyTableName2 = generateTableName
val row2 = spark.sql(s"""call copy_to_table(table=>'$tableName',new_table=>'$copyTableName2',query_type=>'snapshot',as_of_instant=>'$instanceTime')""").collectAsList()
// check exit code
assert(row2.size() == 1 && row2.get(0).get(0) == 0)
val df = spark.sql(s"""select * from $copyTableName2""")
assert(df.count() == 5)
val ids: util.List[Row] = df.selectExpr("id").collectAsList()
assert(ids.containsAll(util.Arrays.asList(Row(1), Row(2), Row(3), Row(4), Row(5))))
}
}
test("Test Call copy_to_table Procedure with incremental") {
withTempDir { tmp =>
val tableName = generateTableName
// create table
spark.sql(
s"""
|create table $tableName (
| id int,
| name string,
| price double,
| ts long
|) using hudi
| location '${tmp.getCanonicalPath}/$tableName'
| tblproperties (
| primaryKey = 'id',
| preCombineField = 'ts'
| )
""".stripMargin)
// insert 4 rows data to table
spark.sql(s"insert into $tableName select 1, 'a1', 10, 1000")
spark.sql(s"insert into $tableName select 2, 'a2', 20, 1500")
// mark beginTime
val beginTime = spark.sql(s"select max(_hoodie_commit_time) from $tableName").collectAsList().get(0).get(0)
spark.sql(s"insert into $tableName select 3, 'a3', 30, 2000")
spark.sql(s"insert into $tableName select 4, 'a4', 40, 2500")
val endTime = spark.sql(s"select max(_hoodie_commit_time) from $tableName").collectAsList().get(0).get(0)
val copyTableName = generateTableName
// Check required fields
checkExceptionContain(s"call copy_to_table(table=>'$tableName',new_table=>'$copyTableName',query_type=>'incremental')")("begin_instance_time and end_instance_time can not be null")
//copy from tableName with begin_instance_time、end_instance_time
val copyCmd = spark.sql(s"call copy_to_table" + s"(table=>'$tableName'" +
s",new_table=>'$copyTableName'" +
s",query_type=>'incremental'" +
s",begin_instance_time=>'$beginTime'" +
s",end_instance_time=>'$endTime')").collectAsList()
assert(copyCmd.size() == 1 && copyCmd.get(0).get(0) == 0)
val df = spark.sql(s"select * from $copyTableName")
assert(df.count() == 2)
val ids = df.selectExpr("id").collectAsList()
assert(ids.containsAll(util.Arrays.asList(Row(3), Row(4))))
}
}
test("Test Call copy_to_table Procedure with read_optimized") {
withTempDir { tmp =>
val tableName = generateTableName
// create mor table with hoodie.compact.inline.max.delta.commits=5
spark.sql(
s"""
|create table $tableName (
| id int,
| name string,
| price double,
| ts long
|) using hudi
| location '${tmp.getCanonicalPath}/$tableName'
| options (
| type='mor',
| primaryKey = 'id',
| preCombineField = 'ts',
| hoodie.compact.inline.max.delta.commits='5',
| hoodie.compact.inline='true'
|
| )
""".stripMargin)
//add 4 delta commit
spark.sql(s"insert into $tableName select 1, 'a1', 10, 1000")
spark.sql(s"update $tableName set ts=2000 where id = 1")
spark.sql(s"update $tableName set ts=3000 where id = 1")
spark.sql(s"update $tableName set ts=4000 where id = 1")
val copyTableName = generateTableName
val copyCmd = spark.sql(s"""call copy_to_table(table=>'$tableName',new_table=>'$copyTableName',query_type=>'read_optimized')""").collectAsList()
assert(copyCmd.size() == 1 && copyCmd.get(0).get(0) == 0)
val copyDf = spark.sql(s"select * from $copyTableName")
assert(copyDf.count() == 1)
//check ts
assert(copyDf.selectExpr("ts").collectAsList().contains(Row(1000)))
// trigger compact (delta_commit==5)
spark.sql(s"update $tableName set ts=5000 where id = 1")
val copyTableName2 = generateTableName
val copyCmd2 = spark.sql(s"""call copy_to_table(table=>'$tableName',new_table=>'$copyTableName2',query_type=>'read_optimized')""").collectAsList()
assert(copyCmd2.size() == 1 && copyCmd2.get(0).get(0) == 0)
val copyDf2 = spark.sql(s"select * from $copyTableName2")
assert(copyDf2.count() == 1)
//check ts
assert(copyDf2.selectExpr("ts").collectAsList().contains(Row(5000)))
}
}
test("Test Call copy_to_table Procedure with append_mode") {
withTempDir { tmp =>
val tableName = generateTableName
// create table
spark.sql(
s"""
|create table $tableName (
| id int,
| name string,
| price double,
| ts long
|) using hudi
| location '${tmp.getCanonicalPath}/$tableName'
| tblproperties (
| primaryKey = 'id',
| preCombineField = 'ts'
| )
""".stripMargin)
// insert data to table
spark.sql(s"insert into $tableName select 1, 'a1', 10, 1000")
spark.sql(s"insert into $tableName select 2, 'a2', 20, 1500")
spark.sql(s"insert into $tableName select 3, 'a3', 30, 2000")
spark.sql(s"insert into $tableName select 4, 'a4', 40, 2500")
val copyTableName = generateTableName
// Check required fields
checkExceptionContain(s"call copy_to_table(table=>'$tableName')")(s"Argument: new_table is required")
val copyCmd = spark.sql(s"call copy_to_table(table=>'$tableName',new_table=>'$copyTableName')").collectAsList()
assert(copyCmd.size() == 1 && copyCmd.get(0).get(0) == 0)
val copyTableCount = spark.sql(s"""select count(1) from $copyTableName""").collectAsList()
assert(copyTableCount.size() == 1 && copyTableCount.get(0).get(0) == 4)
// add 2 rows
spark.sql(s"insert into $tableName select 5, 'a1', 10, 1000")
spark.sql(s"insert into $tableName select 6, 'a2', 20, 1500")
val copyCmd2 = spark.sql(s"call copy_to_table(table=>'$tableName',new_table=>'$copyTableName',save_mode=>'append')").collectAsList()
assert(copyCmd2.size() == 1 && copyCmd2.get(0).get(0) == 0)
val df2 = spark.sql(s"""select * from $copyTableName""")
// total insert 4+6=10 rows
assert(df2.count() == 10)
val ids2 = df2.selectExpr("id").collectAsList()
assert(ids2.containsAll(util.Arrays.asList(Row(1), Row(2), Row(3), Row(4), Row(5), Row(6))))
}
}
test("Test Call copy_to_table Procedure with overwrite_mode") {
withTempDir { tmp =>
val tableName = generateTableName
// create table
spark.sql(
s"""
|create table $tableName (
| id int,
| name string,
| price double,
| ts long
|) using hudi
| location '${tmp.getCanonicalPath}/$tableName'
| tblproperties (
| primaryKey = 'id',
| preCombineField = 'ts'
| )
""".stripMargin)
// insert data to table
spark.sql(s"insert into $tableName select 1, 'a1', 10, 1000")
spark.sql(s"insert into $tableName select 2, 'a2', 20, 1500")
spark.sql(s"insert into $tableName select 3, 'a3', 30, 2000")
spark.sql(s"insert into $tableName select 4, 'a4', 40, 2500")
val copyTableName = generateTableName
// Check required fields
checkExceptionContain(s"call copy_to_table(table=>'$tableName')")(s"Argument: new_table is required")
val copyCmd = spark.sql(s"call copy_to_table(table=>'$tableName',new_table=>'$copyTableName')").collectAsList()
assert(copyCmd.size() == 1 && copyCmd.get(0).get(0) == 0)
val copyTableCount = spark.sql(s"""select count(1) from $copyTableName""").collectAsList()
assert(copyTableCount.size() == 1 && copyTableCount.get(0).get(0) == 4)
// add 2 rows
spark.sql(s"insert into $tableName select 5, 'a1', 10, 1000")
spark.sql(s"insert into $tableName select 6, 'a2', 20, 1500")
val copyCmd2 = spark.sql(s"call copy_to_table(table=>'$tableName',new_table=>'$copyTableName',save_mode=>'Overwrite')").collectAsList()
assert(copyCmd2.size() == 1 && copyCmd2.get(0).get(0) == 0)
val df2 = spark.sql(s"""select * from $copyTableName""")
// total insert 6 rows
assert(df2.count() == 6)
val ids2 = df2.selectExpr("id").collectAsList()
assert(ids2.containsAll(util.Arrays.asList(Row(1), Row(2), Row(3), Row(4), Row(5), Row(6))))
}
}
test("Test Call copy_to_table Procedure with not support mode") {
withTempDir { tmp =>
val tableName = generateTableName
// create table
spark.sql(
s"""
|create table $tableName (
| id int,
| name string,
| price double,
| ts long
|) using hudi
| location '${tmp.getCanonicalPath}/$tableName'
| tblproperties (
| primaryKey = 'id',
| preCombineField = 'ts'
| )
""".stripMargin)
// insert 4 rows data to table
spark.sql(s"insert into $tableName select 1, 'a1', 10, 1000")
spark.sql(s"insert into $tableName select 2, 'a2', 20, 1500")
spark.sql(s"insert into $tableName select 3, 'a3', 30, 2000")
spark.sql(s"insert into $tableName select 4, 'a4', 40, 2500")
val copyTableName = generateTableName
// Check required fields
checkExceptionContain(s"call copy_to_table(table=>'$tableName',new_table=>'$copyTableName',save_mode=>'append1')")(s"save_mode not support append1")
}
}
}