morningman opened a new issue, #16446: URL: https://github.com/apache/doris/issues/16446
[Release Note 1.2.1](https://github.com/apache/doris/issues/15508) Updated: v20230206: d903969f19af4b84af702f47830d909c5fdad65a ## New Features - Multi-Catalog - Support automatic synchronization of Hive metastore - https://doris.apache.org/docs/dev/lakehouse/multi-catalog/#metadata-update - Support reading the Iceberg Snapshot, and viewing the Snapshot history - https://doris.apache.org/en/docs/dev/lakehouse/multi-catalog/iceberg - JDBC Catalog supports PostgreSQL, Clickhouse, Oracle, SQLServer - https://doris.apache.org/en/docs/dev/lakehouse/multi-catalog/jdbc - JDBC Catalog supports Insert operation - TODO: doc - New Functions - `width_bucket`：https://doris.apache.org/docs/dev/sql-manual/sql-functions/width-bucket ## Behavior Changes - Add new session variable `group_by_and_having_use_alias_first`, used to control whether the group and having clauses use alias - https://doris.apache.org/en/docs/dev/advanced/variables ## Improvement - Compaction - Support `Vertical Compaction`. To optimize the compaction overhead and efficiency of wide tables. - TODO: Doc - Support `Segment Compaction`. Fix -238 and -235 issues with high frequency imports. - TODO: Doc - Catalog - Hive Catalog can compatible Hive version 1/2/3 - https://doris.apache.org/docs/dev/lakehouse/multi-catalog/hive#hive-versions - Hive Catalog Support AWS Glue catalog service - TODO: Doc - Hive Catalog can access JuiceFS based HDFS with Broker. - https://doris.apache.org/en/docs/dev/lakehouse/multi-catalog/hive - Iceberg Catalog Support Hive Metastore and Rest Catalog type. - TODO: Doc - ES Catalog support `_id` column mapping。 - https://doris.apache.org/en/docs/dev/lakehouse/multi-catalog/es - Support auto bucket when creating table - TODO: Doc - Optimize Iceberg V2 read performance with large number of delete rows. - Support for reading Iceberg tables after Schema Evolution - Parquet Reader handles column name case correctly. - Support for accessing Hadoop KMS-encrypted HDFS. - https://doris.apache.org/docs/dev/lakehouse/multi-catalog/hive - Support to cancel the Export export task in progress. - https://doris.apache.org/en/docs/dev/sql-manual/sql-reference/Data-Manipulation-Statements/Manipulation/CANCEL-EXPORT - Optimize the performance of `explode_split` with 1x. - Optimize the read performance of nullable columns with 3x - Optimize some problems of Memtracker, improve memory management accuracy, and optimize memory application. ## Bug Fixes - Fixed the possible thread scheduling problem of BE and reduce the `Fragment sent timeout` error caused by BE thread exhaustion #16231 - Fixed various correctness and precision issues of column type `datetimev2`/`decimalv3`. - Fixed the problem data correctness issue with Unique Key Merge-on-Read table. - Fixed various known issues with the Light Schema Change feature. - Fixed various data correctness issues of bitmap type Runtime Filter. - Fixed the problem of poor reading performance of csv reader introduced in version 1.2.1. #15581 - Fixed the problem of BE OOM caused by Spark Load data download phase. #15620 - Fixed possible metadata compatibility issues when upgrading from version 1.1 to version 1.2. #15706 - Fixed the metadata problem when creating JDBC Catalog with Resource. #15692 - Fixed the problem of high CPU usage caused by load operation #16034 - Fixed the problem of FE OOM caused by a large number of failed Broker Load jobs. #15895 - Fixed the problem of precision loss when loading floating-point types #16190 ## Other - Added a metric to view the total rowset and segment numbers on BE - TODO: Doc -- 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]
