neo4j-spark-connector

• First major release of the Neo4j Connector for Apache Spark
• Based around the DataSource API, with polyglot support
• Supports (2.4.5 <= Spark < 3)
• Supports Neo4j 3.5, 4.0, and 4.1

Hello community!
We're happy to share the second pre-release of the Neo4j Connector for Apache Spark!

The connector supports Apache Spark starting from version 2.4.1.

You can find the documentation at the following link.

If you want to provide us feedback, please raise an issue.

In attachment you will find two files:

• neo4j-spark-connector-2.11-4.0.0-pre2.jar which is for Scala 2.11
• neo4j-spark-connector-2.12-4.0.0-pre2.jar which is for Scala 2.12

Hello community!
We're happy to share the first pre-release of the Neo4j Connector for Apache Spark.

The connector supports Apache Spark starting from verson 2.4.1.

You can find the documentation at the following link.

If you want to provide us feedback, please raise an issue.

In attachment you will find two files:

• neo4j-spark-connector-2.11-4.0.0-pre1.jar which is for Scala 2.11
• neo4j-spark-connector-2.12-4.0.0-pre1.jar which is for Scala 2.12

Thanks a lot to @conker84 @nielsdejong @utnaf for your contributions.

Contains the following fixes and improvements.

• Updated saveGraph.updateNodes to use numPartitions instead of batch size.
• Added correct numPartitions calculation for edges
• fixes #133: Neo4jDataFrame#mergeEdgeList should skip node creation (#137)
• fixes #138: Support for Scala 2.12.x (#146)
• fixes #132: Supporting temporal data types using Neo4jDataFrame.mergeEdgeList (#136)
• Updated saveGraph.updateNodes to use numPartitions instead of batch size.

This release makes the spark connector ready for Neo4j 4.0, we also upgraded the spark depedency.

Thanks to @conker84 @moxious and @JakobWyatt-VGW for your work and contributions.

• spark 2.4.5
• neo4j 4.0 and multi-db support
• retry
• separate node from relationship creation
• safe closing of driver and session
• new, shorter config parameters
• improved datetime handling

The release is also available on https://spark-packages.org/package/neo4j-contrib/neo4j-spark-connector

Release with some bugfixes and additions.

Bigger release will be coming in a few weeks.

• Spark 2.4.0

• Neo4j 3.5.x

• Driver 1.7.5

• encryption flag config

• handling of datetime values

• better recursive collection handling

• handling of sequences

• set complex properties on vertices

• documentation for mergeEdgeList with relationship properties

• Neo4jDataFrame#createNodes method and renameMap to mergeEdgeList

thanks to @JakobWyatt-VGW @echohlne @rbramley @tomerlevi1983 @conker84 for your contributions!

This release updates the Neo4j Bolt driver to 1.6.1 and allows for separate read and write transactions in the implementation so that bolt+routing URIs and reading from Read-Replicas should work.

Also updates the spark-dependency to 2.2.1