Apache Hudi

Apache Hudi (pronounced Hoodie) stands for Hadoop Upserts Deletes and Incrementals. Hudi manages the storage of large analytical datasets on DFS (Cloud stores, HDFS or any Hadoop FileSystem compatible storage).

https://hudi.apache.org/

Features

Upsert support with fast, pluggable indexing
Atomically publish data with rollback support
Snapshot isolation between writer & queries
Savepoints for data recovery
Manages file sizes, layout using statistics
Async compaction of row & columnar data
Timeline metadata to track lineage
Optimize data lake layout with clustering

Hudi supports three types of queries:

Snapshot Query - Provides snapshot queries on real-time data, using a combination of columnar & row-based storage (e.g Parquet + Avro).
Incremental Query - Provides a change stream with records inserted or updated after a point in time.
Read Optimized Query - Provides excellent snapshot query performance via purely columnar storage (e.g. Parquet).

Learn more about Hudi at https://hudi.apache.org

Building Apache Hudi from source

Prerequisites for building Apache Hudi:

Unix-like system (like Linux, Mac OS X)
Java 8 (Java 9 or 10 may work)
Git
Maven (>=3.3.1)

# Checkout code and build
git clone https://github.com/apache/hudi.git && cd hudi
mvn clean package -DskipTests

# Start command
spark-2.4.4-bin-hadoop2.7/bin/spark-shell \
  --jars `ls packaging/hudi-spark-bundle/target/hudi-spark-bundle_2.11-*.*.*-SNAPSHOT.jar` \
  --conf 'spark.serializer=org.apache.spark.serializer.KryoSerializer'

To build the Javadoc for all Java and Scala classes:

# Javadoc generated under target/site/apidocs
mvn clean javadoc:aggregate -Pjavadocs

Build with different Spark versions

The default Spark version supported is 2.4.4. To build for different Spark versions and Scala 2.12, use the corresponding profile

Label	Artifact Name for Spark Bundle	Maven Profile Option	Notes
Spark 2.4, Scala 2.11	hudi-spark2.4-bundle_2.11	`-Pspark2.4`	For Spark 2.4.4, which is the same as the default
Spark 2.4, Scala 2.12	hudi-spark2.4-bundle_2.12	`-Pspark2.4,scala-2.12`	For Spark 2.4.4, which is the same as the default and Scala 2.12
Spark 3.1, Scala 2.12	hudi-spark3.1-bundle_2.12	`-Pspark3.1`	For Spark 3.1.x
Spark 3.2, Scala 2.12	hudi-spark3.2-bundle_2.12	`-Pspark3.2`	For Spark 3.2.x
Spark 3, Scala 2.12	hudi-spark3-bundle_2.12	`-Pspark3`	This is the same as `Spark 3.2, Scala 2.12`
Spark, Scala 2.11	hudi-spark-bundle_2.11	Default	The default profile, supporting Spark 2.4.4
Spark, Scala 2.12	hudi-spark-bundle_2.12	`-Pscala-2.12`	The default profile (for Spark 2.4.4) with Scala 2.12

For example,

# Build against Spark 3.2.x (the default build shipped with the public Spark 3 bundle)
mvn clean package -DskipTests -Pspark3.2

# Build against Spark 3.1.x
mvn clean package -DskipTests -Pspark3.1

# Build against Spark 2.4.4 and Scala 2.12
mvn clean package -DskipTests -Pspark2.4,scala-2.12

What about “spark-avro” module?

Starting from versions 0.11, Hudi no longer requires spark-avro to be specified using --packages

Running Tests

Unit tests can be run with maven profile unit-tests.

mvn -Punit-tests test

Functional tests, which are tagged with @Tag("functional"), can be run with maven profile functional-tests.

mvn -Pfunctional-tests test

To run tests with spark event logging enabled, define the Spark event log directory. This allows visualizing test DAG and stages using Spark History Server UI.

mvn -Punit-tests test -DSPARK_EVLOG_DIR=/path/for/spark/event/log

Quickstart

Please visit https://hudi.apache.org/docs/quick-start-guide.html to quickly explore Hudi's capabilities using spark-shell.