release/0.7.9

Features

• #452 - Spark 3.3 support
• #36 - ElasticSearch-hadoop connector support Scala 2.12
• #53 - Add a fallback lineage dispatcher
• #396 - HttpLineageDispatcher to not expect Spline server on the other end

Bugfixes

• #450 - Fix: When Spark is instructed to create a View, Attribute lineage breaks

Tech debts

• #443 - Fix Scala compiler warnings
• #460 - Exclude test classes from code duplication check on Sonar

0.7.8

Features

• #426 Iceberg support

Bugfixes

• Fix handling errors occurred in the ObjectStructureDumper (#427, #137)
• Fix java.lang.NumberFormatException: For input string: "1,146"… (#431)

0.7.7

Bugfixes

• #423 Fix "json4s-ext" incompatibility for Spark 3.2

0.7.6 - Spark 3.2 support

Features

• #353, #419 - Add Spark 3.2 support

Bugfixes

• #410 - fix compilation on case-insensitive file systems

0.7.4

Bugfixes

• #408 - fix ObjectStructureDumper exception logging
• #417 - fix timeout settings not propagated to the server

0.7.3

Bugfixes

#383 - fix Kafka dispatcher error Cannot perform operation after producer has been closed

0.7.2

Bugfixes

#370 - Fix Spark/Hadoop configuration string array properties parsing

0.7.1

Bugfixes

#346 - Unexpected error using spline agent 0.7 with AWS Glue 3.0

Improvements

#348 - Capture more details about object data types

0.7.0

Summary

This release includes two important improvements:

1. The agent dependencies have been optimized. It resulted in 4 times decrease of the bundle JAR size - 3 Mb instead of 12 Mb (#91)
2. Switching to UUID version 5 in combination with a few other changes allowed to make the agent output deterministic.It helps deduplicating lineage of recurring jobs which reduces the Spline database size and load on the server. (#323)

---

Bugfixes

#314 - KafkaDispatcher: set max.in.flight.requests.per.connection to 1 by default to avoid reordering
#328 - Lineage harvester output is not deterministic

Improvements

#91 - Optimize dependencies
#323 - Support for multiple types of ID
#326 - Logging: Spline init errors in BEST_EFFORT mode should be logged as WARN

Documentation

#297 - Document versioning strategy

0.6.4 - Java 11+ & Google BigQuery

Features

• #277 - Google BigQuery support (Thanks @mohanajuhi166)

Bugfixes

• #308 - Fix an error when running REPL under Java 11+