[
https://issues.apache.org/jira/browse/HUDI-7016?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Ethan Guo updated HUDI-7016:
----------------------------
Description:
When running Spark quickstart with deletes using the Hudi Spark bundle and
turning on position encoding for deletes and updates
("hoodie.write.record.positions=true"), the write fails with the following
exception:
{code:java}
Caused by: java.lang.NoSuchMethodError:
org.roaringbitmap.longlong.Roaring64NavigableMap.serializePortable(Ljava/io/DataOutput;)V
at
org.apache.hudi.common.table.log.LogReaderUtils.encodePositions(LogReaderUtils.java:128)
at
org.apache.hudi.common.table.log.LogReaderUtils.encodePositions(LogReaderUtils.java:109)
at
org.apache.hudi.common.table.log.block.HoodieLogBlock.addRecordPositionsToHeader(HoodieLogBlock.java:147)
at
org.apache.hudi.common.table.log.block.HoodieDeleteBlock.<init>(HoodieDeleteBlock.java:88)
at
org.apache.hudi.io.HoodieAppendHandle.appendDataAndDeleteBlocks(HoodieAppendHandle.java:480)
at
org.apache.hudi.io.HoodieAppendHandle.doAppend(HoodieAppendHandle.java:454)
at
org.apache.hudi.table.action.deltacommit.BaseSparkDeltaCommitActionExecutor.handleUpdate(BaseSparkDeltaCommitActionExecutor.java:83)
at
org.apache.hudi.table.action.commit.BaseSparkCommitActionExecutor.handleUpsertPartition(BaseSparkCommitActionExecutor.java:313)
... 28 more {code}
Spark quickstart run:
{code:java}
// spark-shell
import org.apache.hudi.QuickstartUtils._
import scala.collection.JavaConversions._
import org.apache.spark.sql.SaveMode._
import org.apache.hudi.DataSourceReadOptions._
import org.apache.hudi.DataSourceWriteOptions._
import org.apache.hudi.config.HoodieWriteConfig._
import org.apache.hudi.common.model.HoodieRecord
val tableName = "hudi_trips_cow"
val basePath = "file:///tmp/hudi_trips_mor"
val dataGen = new DataGenerator
val inserts = convertToStringList(dataGen.generateInserts(1000))
val df = spark.read.json(spark.sparkContext.parallelize(inserts, 2))
df.write.format("hudi").
option(PRECOMBINE_FIELD_OPT_KEY, "ts").
option(RECORDKEY_FIELD_OPT_KEY, "uuid").
option(PARTITIONPATH_FIELD_OPT_KEY, "partitionpath").
option("hoodie.logfile.data.block.format", "parquet").
option("hoodie.write.record.positions", "true").
option("hoodie.datasource.write.table.type", "MERGE_ON_READ").
option(TABLE_NAME, tableName + "true").
mode(Overwrite).
save(basePath + "true")
df.write.format("hudi").
option(PRECOMBINE_FIELD_OPT_KEY, "ts").
option(RECORDKEY_FIELD_OPT_KEY, "uuid").
option(PARTITIONPATH_FIELD_OPT_KEY, "partitionpath").
option("hoodie.logfile.data.block.format", "parquet").
option("hoodie.write.record.positions", "false").
option("hoodie.datasource.write.table.type", "MERGE_ON_READ").
option(TABLE_NAME, tableName + "false").
mode(Overwrite).
save(basePath + "false")
spark.read.format("hudi").load(basePath +
"true").createOrReplaceTempView("hudi_trips_snapshot")
val ds = spark.sql("select uuid, partitionpath from
hudi_trips_snapshot").sample(0.01)
// issue deletes
val deletes = dataGen.generateDeletes(ds.collectAsList())
val hardDeleteDf = spark.read.json(spark.sparkContext.parallelize(deletes, 2))
hardDeleteDf.write.format("hudi").
option(OPERATION_OPT_KEY, "delete").
option(PRECOMBINE_FIELD_OPT_KEY, "ts").
option(RECORDKEY_FIELD_OPT_KEY, "uuid").
option(PARTITIONPATH_FIELD_OPT_KEY, "partitionpath").
option("hoodie.logfile.data.block.format", "parquet").
option("hoodie.write.record.positions", "true").
option("hoodie.datasource.write.table.type", "MERGE_ON_READ").
option(TABLE_NAME, tableName + "true").
mode(Append).
save(basePath + "true")
hardDeleteDf.write.format("hudi").
option(OPERATION_OPT_KEY, "delete").
option(PRECOMBINE_FIELD_OPT_KEY, "ts").
option(RECORDKEY_FIELD_OPT_KEY, "uuid").
option(PARTITIONPATH_FIELD_OPT_KEY, "partitionpath").
option("hoodie.logfile.data.block.format", "parquet").
option("hoodie.write.record.positions", "false").
option("hoodie.datasource.write.table.type", "MERGE_ON_READ").
option(TABLE_NAME, tableName + "false").
mode(Append).
save(basePath + "false")
spark.read.format("hudi").
option("hoodie.datasource.read.use.new.parquet.file.format", "true").
option("hoodie.file.group.reader.enabled", "true").
load(basePath + "true").count()
spark.read.format("hudi").
option("hoodie.datasource.read.use.new.parquet.file.format", "true").
option("hoodie.file.group.reader.enabled", "true").
load(basePath + "false").count() {code}
> Fix bundling of RoaringBitmap dependency
> ----------------------------------------
>
> Key: HUDI-7016
> URL: https://issues.apache.org/jira/browse/HUDI-7016
> Project: Apache Hudi
> Issue Type: Bug
> Reporter: Ethan Guo
> Assignee: Ethan Guo
> Priority: Blocker
> Fix For: 1.0.0
>
>
> When running Spark quickstart with deletes using the Hudi Spark bundle and
> turning on position encoding for deletes and updates
> ("hoodie.write.record.positions=true"), the write fails with the following
> exception:
> {code:java}
> Caused by: java.lang.NoSuchMethodError:
> org.roaringbitmap.longlong.Roaring64NavigableMap.serializePortable(Ljava/io/DataOutput;)V
> at
> org.apache.hudi.common.table.log.LogReaderUtils.encodePositions(LogReaderUtils.java:128)
> at
> org.apache.hudi.common.table.log.LogReaderUtils.encodePositions(LogReaderUtils.java:109)
> at
> org.apache.hudi.common.table.log.block.HoodieLogBlock.addRecordPositionsToHeader(HoodieLogBlock.java:147)
> at
> org.apache.hudi.common.table.log.block.HoodieDeleteBlock.<init>(HoodieDeleteBlock.java:88)
> at
> org.apache.hudi.io.HoodieAppendHandle.appendDataAndDeleteBlocks(HoodieAppendHandle.java:480)
> at
> org.apache.hudi.io.HoodieAppendHandle.doAppend(HoodieAppendHandle.java:454)
> at
> org.apache.hudi.table.action.deltacommit.BaseSparkDeltaCommitActionExecutor.handleUpdate(BaseSparkDeltaCommitActionExecutor.java:83)
> at
> org.apache.hudi.table.action.commit.BaseSparkCommitActionExecutor.handleUpsertPartition(BaseSparkCommitActionExecutor.java:313)
> ... 28 more {code}
> Spark quickstart run:
>
> {code:java}
> // spark-shell
> import org.apache.hudi.QuickstartUtils._
> import scala.collection.JavaConversions._
> import org.apache.spark.sql.SaveMode._
> import org.apache.hudi.DataSourceReadOptions._
> import org.apache.hudi.DataSourceWriteOptions._
> import org.apache.hudi.config.HoodieWriteConfig._
> import org.apache.hudi.common.model.HoodieRecord
> val tableName = "hudi_trips_cow"
> val basePath = "file:///tmp/hudi_trips_mor"
> val dataGen = new DataGenerator
> val inserts = convertToStringList(dataGen.generateInserts(1000))
> val df = spark.read.json(spark.sparkContext.parallelize(inserts, 2))
> df.write.format("hudi").
> option(PRECOMBINE_FIELD_OPT_KEY, "ts").
> option(RECORDKEY_FIELD_OPT_KEY, "uuid").
> option(PARTITIONPATH_FIELD_OPT_KEY, "partitionpath").
> option("hoodie.logfile.data.block.format", "parquet").
> option("hoodie.write.record.positions", "true").
> option("hoodie.datasource.write.table.type", "MERGE_ON_READ").
> option(TABLE_NAME, tableName + "true").
> mode(Overwrite).
> save(basePath + "true")
> df.write.format("hudi").
> option(PRECOMBINE_FIELD_OPT_KEY, "ts").
> option(RECORDKEY_FIELD_OPT_KEY, "uuid").
> option(PARTITIONPATH_FIELD_OPT_KEY, "partitionpath").
> option("hoodie.logfile.data.block.format", "parquet").
> option("hoodie.write.record.positions", "false").
> option("hoodie.datasource.write.table.type", "MERGE_ON_READ").
> option(TABLE_NAME, tableName + "false").
> mode(Overwrite).
> save(basePath + "false")
> spark.read.format("hudi").load(basePath +
> "true").createOrReplaceTempView("hudi_trips_snapshot")
> val ds = spark.sql("select uuid, partitionpath from
> hudi_trips_snapshot").sample(0.01)
> // issue deletes
> val deletes = dataGen.generateDeletes(ds.collectAsList())
> val hardDeleteDf = spark.read.json(spark.sparkContext.parallelize(deletes, 2))
> hardDeleteDf.write.format("hudi").
> option(OPERATION_OPT_KEY, "delete").
> option(PRECOMBINE_FIELD_OPT_KEY, "ts").
> option(RECORDKEY_FIELD_OPT_KEY, "uuid").
> option(PARTITIONPATH_FIELD_OPT_KEY, "partitionpath").
> option("hoodie.logfile.data.block.format", "parquet").
> option("hoodie.write.record.positions", "true").
> option("hoodie.datasource.write.table.type", "MERGE_ON_READ").
> option(TABLE_NAME, tableName + "true").
> mode(Append).
> save(basePath + "true")
> hardDeleteDf.write.format("hudi").
> option(OPERATION_OPT_KEY, "delete").
> option(PRECOMBINE_FIELD_OPT_KEY, "ts").
> option(RECORDKEY_FIELD_OPT_KEY, "uuid").
> option(PARTITIONPATH_FIELD_OPT_KEY, "partitionpath").
> option("hoodie.logfile.data.block.format", "parquet").
> option("hoodie.write.record.positions", "false").
> option("hoodie.datasource.write.table.type", "MERGE_ON_READ").
> option(TABLE_NAME, tableName + "false").
> mode(Append).
> save(basePath + "false")
> spark.read.format("hudi").
> option("hoodie.datasource.read.use.new.parquet.file.format", "true").
> option("hoodie.file.group.reader.enabled", "true").
> load(basePath + "true").count()
> spark.read.format("hudi").
> option("hoodie.datasource.read.use.new.parquet.file.format", "true").
> option("hoodie.file.group.reader.enabled", "true").
> load(basePath + "false").count() {code}
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
