-
Notifications
You must be signed in to change notification settings - Fork 3
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
- Loading branch information
Showing
4 changed files
with
541 additions
and
4 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
227 changes: 227 additions & 0 deletions
227
pramen/core/src/main/scala/za/co/absa/pramen/core/sink/SparkSink.scala
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,227 @@ | ||
| /* | ||
| * Copyright 2022 ABSA Group Limited | ||
| * | ||
| * Licensed 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 za.co.absa.pramen.core.sink | ||
|
|
||
| import com.typesafe.config.Config | ||
| import org.apache.hadoop.fs.Path | ||
| import org.apache.spark.sql.{DataFrame, SparkSession} | ||
| import org.slf4j.LoggerFactory | ||
| import za.co.absa.pramen.api.{ExternalChannelFactory, MetastoreReader, Sink} | ||
| import za.co.absa.pramen.core.config.Keys.KEYS_TO_REDACT | ||
| import za.co.absa.pramen.core.utils.ConfigUtils | ||
|
|
||
| import java.time.LocalDate | ||
|
|
||
| /** | ||
| * This sink allows writing data using Spark, similarly as you would do using 'df.write.format(...).save(...)'. | ||
| * | ||
| * In order to use the sink you need to define sink parameters. | ||
| * | ||
| * Example sink definition: | ||
| * {{{ | ||
| * { | ||
| * # Define a name to reference from the pipeline: | ||
| * name = "spark_sink" | ||
| * factory.class = "za.co.absa.pramen.core.sink.SparkSink" | ||
| * | ||
| * # Output format. Can be: csv, parquet, json, delta, etc (anything supported by Spark). Default: parquet | ||
| * format = "parquet" | ||
| * | ||
| * # Save mode. Can be overwrite, append, ignore, errorifexists. Default: errorifexists | ||
| * mode = "overwrite" | ||
| * | ||
| * ## Only one of these following two options should be specified | ||
| * # Optionally repartition the dataframe according to the specified number of partitions | ||
| * number.of.partitions = 10 | ||
| * # Optionally repartition te dataframe according to the number of records per partition | ||
| * records.per.partition = 1000000 | ||
| * | ||
| * # If true (default), the data will be saved even if it does not contain any records. If false, the saving will be skipped | ||
| * save.empty = true | ||
| * | ||
| * # If non-empty, the data will be partitioned by the specified columns at the output path. Default: [] | ||
| * partition.by = [ pramen_info_date ] | ||
| * | ||
| * # These are additional option passed to the writer as 'df.write(...).options(...) | ||
| * option { | ||
| * compression = "gzip" | ||
| * } | ||
| * } | ||
| * }}} | ||
| * | ||
| * Here is an example of a sink definition in a pipeline. As for any other operation you can specify | ||
| * dependencies, transformations, filters and columns to select. | ||
| * | ||
| * {{{ | ||
| * { | ||
| * name = "Spark sink" | ||
| * type = "sink" | ||
| * sink = "spark_sink" | ||
| * | ||
| * schedule.type = "daily" | ||
| * | ||
| * # Optional dependencies | ||
| * dependencies = [ | ||
| * { | ||
| * tables = [ dependent_table ] | ||
| * date.from = "@infoDate" | ||
| * } | ||
| * ] | ||
| * | ||
| * tables = [ | ||
| * { | ||
| * input.metastore.table = metastore_table | ||
| * output.path = "/datalake/base/path" | ||
| * | ||
| * # Date range to read the source table for. By default the job information date is used. | ||
| * # But you can define an arbitrary expression based on the information date. | ||
| * # More: see the section of documentation regarding date expressions, an the list of functions allowed. | ||
| * date { | ||
| * from = "@infoDate" | ||
| * to = "@infoDate" | ||
| * } | ||
| * | ||
| * transformations = [ | ||
| * { col = "col1", expr = "lower(some_string_column)" } | ||
| * ], | ||
| * filters = [ | ||
| * "some_numeric_column > 100" | ||
| * ] | ||
| * columns = [ "col1", "col2", "col2", "some_numeric_column" ] | ||
| * | ||
| * # This overrides options of the sink | ||
| * sink { | ||
| * mode = "append" | ||
| * | ||
| * # These are additional options passed to the writer as 'df.write(...).options(...)' | ||
| * option { | ||
| * compression = "snappy" | ||
| * } | ||
| * } | ||
| * } | ||
| * ] | ||
| * } | ||
| * }}} | ||
| * | ||
| */ | ||
| class SparkSink(format: String, | ||
| formatOptions: Map[String, String], | ||
| mode: String, | ||
| partitionBy: Seq[String], | ||
| numberOfPartitions: Option[Int], | ||
| recordsPerPartition: Option[Long], | ||
| saveEmpty: Boolean, | ||
| sinkConfig: Config) extends Sink { | ||
|
|
||
| import za.co.absa.pramen.core.sink.SparkSink._ | ||
|
|
||
| private val log = LoggerFactory.getLogger(this.getClass) | ||
|
|
||
| override val config: Config = sinkConfig | ||
|
|
||
| override def connect(): Unit = {} | ||
|
|
||
| override def close(): Unit = {} | ||
|
|
||
| override def send(df: DataFrame, | ||
| tableName: String, | ||
| metastore: MetastoreReader, | ||
| infoDate: LocalDate, | ||
| options: Map[String, String]) | ||
| (implicit spark: SparkSession): Long = { | ||
| val outputPath = getOutputPath(tableName, options) | ||
| val recordCount = df.count() | ||
|
|
||
| if (recordCount > 0 || saveEmpty) { | ||
| log.info(s"Saving $recordCount records to folder: ${outputPath.toUri.toString}") | ||
| log.info(s"Options passed for '$format':") | ||
| ConfigUtils.renderExtraOptions(formatOptions, KEYS_TO_REDACT)(log.info) | ||
|
|
||
| val dfToWrite = applyRepartitioning(df, recordCount, tableName) | ||
| writeData(dfToWrite, outputPath) | ||
| } else { | ||
| log.info(s"Nothing to save to folder: ${outputPath.toUri.toString}") | ||
| } | ||
|
|
||
| recordCount | ||
| } | ||
|
|
||
| private[core] def writeData(df: DataFrame, outputPath: Path): Unit = { | ||
| df | ||
| .write | ||
| .partitionBy(partitionBy: _*) | ||
| .format(format) | ||
| .mode(mode) | ||
| .options(formatOptions) | ||
| .save(outputPath.toUri.toString) | ||
| } | ||
|
|
||
| private[core] def applyRepartitioning(df: DataFrame, recordCount: Long, tableName: String): DataFrame = { | ||
| (numberOfPartitions, recordsPerPartition) match { | ||
| case (Some(_), Some(_)) => | ||
| throw new IllegalArgumentException( | ||
| s"Both $NUMBER_OF_PARTITIONS_KEY and $RECORDS_PER_PARTITION_KEY are specified for Spark sink," + | ||
| s"table: $tableName. Please specify only one of those options") | ||
| case (Some(nop), None) => | ||
| log.info(s"Repartitioning to $nop partitions") | ||
| df.repartition(nop) | ||
| case (None, Some(rpp)) => | ||
| val n = Math.max(1, Math.ceil(recordCount.toDouble / rpp)).toInt | ||
| log.info(s"Repartitioning to $n partitions") | ||
| df.repartition(n) | ||
| case (None, None) => | ||
| df | ||
| } | ||
| } | ||
|
|
||
| private[core] def getOutputPath(tableName: String, options: Map[String, String]): Path = { | ||
| if (!options.contains(OUTPUT_PATH_KEY)) { | ||
| throw new IllegalArgumentException(s"$OUTPUT_PATH_KEY is not specified for Spark sink, table: $tableName") | ||
| } | ||
|
|
||
| new Path(options(OUTPUT_PATH_KEY)) | ||
| } | ||
|
|
||
| } | ||
|
|
||
| object SparkSink extends ExternalChannelFactory[SparkSink] { | ||
| val OUTPUT_PATH_KEY = "path" | ||
|
|
||
| val FORMAT_KEY = "format" | ||
| val MODE_KEY = "mode" | ||
| val PARTITION_BY_KEY = "partition.by" | ||
| val NUMBER_OF_PARTITIONS_KEY = "number.of.partitions" | ||
| val RECORDS_PER_PARTITION_KEY = "records.per.partition" | ||
| val SAVE_EMPTY_KEY = "save.empty" | ||
|
|
||
| val DEFAULT_FORMAT = "parquet" | ||
| val DEFAULT_MODE = "errorifexists" | ||
| val DEFAULT_SAVE_EMPTY = true | ||
|
|
||
| override def apply(conf: Config, parentPath: String, spark: SparkSession): SparkSink = { | ||
| new SparkSink( | ||
| ConfigUtils.getOptionString(conf, FORMAT_KEY).getOrElse(DEFAULT_FORMAT), | ||
| ConfigUtils.getExtraOptions(conf, "option"), | ||
| ConfigUtils.getOptionString(conf, MODE_KEY).getOrElse(DEFAULT_MODE), | ||
| ConfigUtils.getOptListStrings(conf, PARTITION_BY_KEY), | ||
| ConfigUtils.getOptionInt(conf, NUMBER_OF_PARTITIONS_KEY), | ||
| ConfigUtils.getOptionLong(conf, RECORDS_PER_PARTITION_KEY), | ||
| ConfigUtils.getOptionBoolean(conf, SAVE_EMPTY_KEY).getOrElse(DEFAULT_SAVE_EMPTY), | ||
| conf | ||
| ) | ||
| } | ||
| } |
Oops, something went wrong.