007vedant opened a new issue, #10812: URL: https://github.com/apache/hudi/issues/10812
**_Tips before filing an issue_** - Have you gone through our [FAQs](https://hudi.apache.org/learn/faq/)? - Join the mailing list to engage in conversations and get faster support at [email protected]. - If you have triaged this as a bug, then file an [issue](https://issues.apache.org/jira/projects/HUDI/issues) directly. **Describe the problem you faced** I'm having a Copy-On-Write (COW) table in AWS S3 bucket with some date based partitions. When I perform a `delete_partitions` operation on the Hudi table with cleaner policy configuration of `KEEP_LATEST_FILE_VERSIONS = 1`, hudi not only removes the partitions from the metadata (expected) but also actually deletes the partitions and the base parquet files from the S3 storage (unexpected). I'm unable to understand why this is happening. As per Hudi's latest documentation, `KEEP_LATEST_FILE_VERSIONS` cleaner configurations deletes all but the latest file version from the storage. So when I perform the `delete_partitions` operation, are all the file groups in those partitions marked with a special `to_be_deleted` kind of version or the cleaning action first interacts with the table metadata, finds that the partitions and their files groups have been dropped and thus performs the actual cleaner action over those partitions. I'm more curious on why this is happening rather than how is this happening, as this is the favorable outcome in my use case. NOTE: The actual file eviction from storage was not happening when the cleaner configuration was `KEEP_LATEST_COMMITS = 1`. In this is case, I had to perform an additional commit to trigger the cleaner action. Write configurations used: ``` val FILE_VERSIONS_TO_RETAIN = 1 val options = Map[String, String]( DataSourceWriteOptions.OPERATION.key -> DataSourceWriteOptions.DELETE_PARTITION_OPERATION_OPT_VAL, DataSourceWriteOptions.PARTITIONS_TO_DELETE.key -> listOfPartitions, HoodieCleanConfig.CLEANER_POLICY.key -> HoodieCleaningPolicy.KEEP_LATEST_FILE_VERSIONS.name, HoodieCleanConfig.CLEANER_FILE_VERSIONS_RETAINED.key -> FILE_VERSIONS_TO_RETAIN ) // empty df is an empty dataframe emptyDf.write .format("hudi") .options(cleanerOptions) .mode("append") .save(tablePath) ``` A clear and concise description of the problem. **To Reproduce** Steps to reproduce the behavior: 1. Create a Copy-On-Write (COW) hudi table in AWS S3. 2. Write some data into the table partitioned using `hoodie.datasource.write.partitionpath.field` 3. Drop some partitions using the write configurations provided above. **Expected behavior** A clear and concise description of what you expected to happen. **Environment Description** * Hudi version : hudi-spark3.3-bundle_2.12 * Spark version : 3.3.1 * Hive version : 3.1.3 * Hadoop version : 3.3.3 * Storage (HDFS/S3/GCS..) : S3 * Running on Docker? (yes/no) : no (AWS EMR 6.10.0) **Additional context** Add any other context about the problem here. **Stacktrace** ```Add the stacktrace of the error.``` -- 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]
