andygrove commented on PR #3500: URL: https://github.com/apache/datafusion-comet/pull/3500#issuecomment-3892624872
## CI Failure Analysis: PR #3500 vs base df52 branch (#3470) I compared the CI failures on this PR against the base `df52` branch (run [21949559770](https://github.com/apache/datafusion-comet/actions/runs/21949559770)) to identify which failures are new vs pre-existing. ### New failure introduced by this PR | Test Suite | Jobs Affected | Notes | |---|---|---| | **`ParquetFieldIdIOSuite`** | `sql_core-1` on Spark 3.4.3, 3.5.8, 4.0.1 | Not present on base `df52` branch. This is the only new test failure introduced by this PR. | ### Pre-existing failures (same on base df52 branch) | Test Suite | Jobs Affected | Notes | |---|---|---| | `ParquetSchemaSuite` | `sql_core-1` on all Spark versions, `native_datafusion-sql_core-1/spark-3.5.8` | Same failures on base branch | | `SQLQueryTestSuite` | `sql_core-2` on Spark 3.5.8 (auto + native_datafusion), Spark 4.0.1 (auto) | Same failures on base branch | | `InsertSuite` | `native_datafusion-sql_core-1/spark-3.5.8` | Same failure on base branch | | `width_bucket - with range data` | All `[expressions]` jobs (Linux + macOS, all Spark versions) | Same failure on base branch | | `select column with default value` | `native_datafusion [fuzz]` | Pre-existing, tracked in #3482 | | Miri build failure | All branches | Systemic toolchain issue, tracked in #3499, fix in #3504 | ### Resolved by this PR | Test Suite | Notes | |---|---| | `ParquetVariantShreddingSuite` | Was failing on base `df52` branch for Spark 4.0.1. This PR (which includes the skip) eliminates this failure. | ### No longer failing on this PR (compared to base) | Test Suite | Notes | |---|---| | `ParquetV1SchemaPruningSuite` | Was failing on base branch in `sql_core-2` for Spark 3.4.3, 3.5.8, and 4.0.1. Not failing in this PR — likely fixed by the schema pruning changes. | ### Summary The only **new regression** is `ParquetFieldIdIOSuite` (failing across all 3 Spark versions in `sql_core-1`). Everything else is either pre-existing on the `df52` branch or actually improved by this PR (fixing `ParquetV1SchemaPruningSuite` and skipping `ParquetVariantShreddingSuite`). -- 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]
