[SPARK-50112] Moving Avro files to sql/core so they can be used by TransformWithState operator

• Type: New Feature
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 4.0.0
• Fix Version/s: 4.0.0
• Component/s: Structured Streaming
• Labels: pull-request-available
• Assignee: Eric Marnadi
• Reporter: Eric Marnadi
• Created: 25/Oct/24 03:46
• Updated: 06/Nov/24 03:21
• Resolved: 05/Nov/24 23:00

Description

With the introduction of the TransformWithState operator, we are going to persist all state information in the StateStore with Avro encoding, instead of UnsafeRow. This is because UnsafeRow is not backwards compatible and can change between Spark releases, potentially causing StateStore corruptions. Avro is backwards-compatible and can be reliably used. This change will move the necessary files to the sql/core directory so Avro encoding can be used by stateful streaming operators.

Issue Links

• links to: GitHub Pull Request #48650