June 2026 monthly summary for the apache/fluss repository focused on improving observability and data quality through targeted bug fixes in the scanner metrics reporting. The patch corrected a typo in the scanner metric name, preventing misreported metrics and enabling reliable dashboards and alerts. This work reduces monitoring drift and supports data-driven decision making across product and operations teams.

February 2026 monthly summary for apache/iceberg-rust: Implemented predicate pushdown of isnan() from DataFusion into Iceberg’s IsNan/NotNan predicates, enabling more efficient file pruning via manifest nan_value_counts for float/double columns. Introduced a runtime translator (scalar_function_to_iceberg_predicate) to map isnan(col) to Iceberg’s Unary(IsNan, col) and added automatic negation handling for NOT isnan(col) through the existing rewrite_not path. Added tests validating pushdown correctness and pruning behavior. This work tightens DataFusion-Iceberg integration, reduces IO, and aligns with Iceberg manifest pruning spec. Commits include 4f5274a8afd835e746167e48683207b71b20115a (feat(datafusion): support isnan predicate pushdown to Iceberg).