malinjawi opened a new issue, #12912:
URL: https://github.com/apache/gluten/issues/12912
### Backend
VL (Velox)
### Bug description
After #12823, Gluten automatically enables `IcebergSparkSessionExtensions`
whenever Iceberg is on the classpath.
This affects every Spark session, including non-Iceberg DataSource V2
tables. With Spark 3.5 and Iceberg 1.10.0, Iceberg's row-lineage rules can
throw a `MatchError` during `UPDATE` and `MERGE`. The Iceberg parser can also
unexpectedly wrap other configured parsers.
Expected behavior: Iceberg extensions should only be enabled when explicitly
configured by an Iceberg application.
Suggested fix:
1. Stop registering the Iceberg extension globally.
2. Configure it explicitly in Iceberg tests and applications.
3. Add the Iceberg runtime to the `gluten-ut` Iceberg profile.
Community CI missed this because the Iceberg runtime was not on the
`gluten-ut` test classpath, so the Iceberg component remained inactive.
### Gluten version
main branch
### Spark version
Spark-3.5.x
### Spark configurations
spark.plugins=org.apache.gluten.GlutenPlugin
Iceberg runtime 1.10.0 is present on the classpath
### System information
_No response_
### Relevant logs
```bash
scala.MatchError:
org.apache.spark.sql.connector.catalog.InMemoryRowLevelOperationTable
at
org.apache.spark.sql.catalyst.analysis.RewriteOperationForRowLineage.shouldUpdatePlan
at
org.apache.spark.sql.catalyst.analysis.RewriteMergeIntoTableForRowLineage.apply
```
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
