cloud-fan commented on PR #59728: URL: https://github.com/apache/spark/pull/59728#issuecomment-6371577273
A follow-up note, not blocking this PR since the issue predates it: `ALTER TABLE .. ADD IF NOT EXISTS PARTITION` can still overwrite an existing partition through the stats branch of `AlterTableAddPartitionCommand`: https://github.com/apache/spark/blob/7d16a2ba4afdf64731aa75f77b1dac54e07686be/sql/core/src/main/scala/org/apache/spark/sql/execution/command/ddl.scala#L624-L632 When the table has stats, `spark.sql.statistics.size.autoUpdate.enabled` is true, and the requested locations are non-empty, the command calls `catalog.alterPartitions(table.identifier, parts)` with *all* requested partitions. So an existing partition named in the command still gets its storage (e.g. the `LOCATION`) replaced, even though `createPartitions` now skips it. `addedSize` also counts the requested locations of existing partitions, which inflates `sizeInBytes`. This is at the command level, so it applies to both the in-memory and Hive catalogs. Could we fix this in a follow-up by excluding the already-existing specs from both `addedSize` and the `alterPartitions` call? -- 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]
