PolarDB:PolarDB for MySQL 8.0.2 release notes

Description

• Added database- and table-level Branch/FORK capability. You can use SQL statements to create writable branches from existing tables or databases within seconds without copying user data. Branches read a consistent snapshot of the source table at creation time and use page-level Copy-on-Write isolation, so writes do not affect the source table.

• Added parallel archiving for partitioned tables. You can concurrently archive partitions that use the InnoDB engine as X-Engine, ORC, or CSV storage engine partitions in hybrid partitioned tables, or use part2table to concurrently archive InnoDB partitions as OSS tables. Multi-threaded archiving improves performance when an individual partition contains a large data volume.

• Added support for parallel partition archiving: InnoDB engine partitions can be archived in parallel to X-Engine, ORC, or CSV engine partitions, forming a hybrid partitioned table. The part2table command is also supported to archive InnoDB partitions in parallel to OSS tables. Parallel archiving uses multiple threads simultaneously, significantly improving speed for partitions with large data volumes.

• Added support for defining secondary indexes on materialized views, and for recovering, importing, and exporting materialized view refreshes.

• Added Multi-master cluster (Limitless) transparent distributed SELECT ... LOCK IN SHARE MODE distributed query capability, and ensuring SHARE row lock transaction lifecycle semantics.

• Added BM25 scoring, phrase queries, combined queries, and synonym support for In-Memory Column Index (IMCI) full-text indexes.

• Added multi-worker parallel execution for distributed DML and DQL on every read-write node in Multi-master cluster (Limitless) Edition. Set max_parallel_degree in the session to enable this capability.

• X-Engine column storing partitioned table support LOAD DATA bypass import (direct load), each partition independently builds columnar data and two-phase unified commit, import failure zero residue; Previously partitioned table could only using regular import path.

• Multi-master cluster distributed DML, DQL support multiple worker parallel, each read/write node (RW) all support, significantly improving execution speed. need in the session specifying max_parallel_degree take effect.

• Added cross-node query-result transfer optimization and improved connection stability in exceptional scenarios.

• Added binlog cache support for read-only nodes, reducing IOPS overhead from concurrent binlog reads.

• Added support for Online Copy PARALLEL DDL: InnoDB tables can be converted online to X-Engine tables using the ALTER TABLE ... ENGINE=xengine statement, with concurrent DML access allowed during conversion without service interruption. ALTER TABLE ... ENGINE=orc/csv also supports parallel COPY DDL.

• Added an X-Engine parameter that controls whether the default database character set collation is used.

• Added GDN multi-write test cases for In-Memory Column Index (IMCI) functionality.

• X-Engine partitioned tables now support partition-level conversion from row store format to columnar format. Syntax: ALTER TABLE table_name CHANGE PARTITION partition_name ENGINE=xengine DATA_FORMAT=column;.

• Added IBD OSS support for partitioned tables, enabling hybrid partitions and expanding OSS support.

• Added flush status support to the hot rows optimization feature, allowing incremental changes to performance statistics to be observed during testing.

• The hot row optimization feature (hotspot) now supports FLUSH STATUS to refresh statistics, making it convenient to observe incremental changes during testing.

• Added innodb_polar_max_delay_time_in_second parameter, support primary node based on redo apply latency threshold proactively evict physical read-only node.

• PolarDB Orca support Redis protocol monitor command, returns Orca processing all command.

• Added Redis-protocol MONITOR command support to PolarDB Orca, which returns all commands processed by Orca.

Description

• Optimized memory-allocation strategies for X-Engine, CTable, and In-Memory Column Index (IMCI) across different operating modes and high-availability failover scenarios.

• Optimized materialized view incremental refresh, ignoring not affecting materialized view results base table UPDATE, avoiding invalid JOIN computation and container table writes.

• Optimized In-Memory Column Index (IMCI) deep paging queries by adding the QuickSelect TopK execution mode to replace full sorting and reduce sorting overhead for large offsets and multi-row scenarios.

• Optimized information_schema.INNODB_LOG_INFO memory usage, avoiding in redo log accumulation in scenarios where occurring additional overhead.

• Optimized binlog file switching time and binlog I/O efficiency, preventing transaction commits from being blocked when many binlog files exist.

• Optimized the distributed memory pool and added related parameters and monitoring metrics.

Description

• Fixed Multi-master cluster remote_exec_cmd in prepared statement during repeated execution, could cause server process crash.

• Fixed materialized view MATERIALIZED OPTIONS in engine=xengine in X-Engine unavailable when was not validated, causing background refresh to crash.

• Fixed an issue where get_status directly wrote to serverless.info during Serverless Multi-master cluster forward meta operations, causing file I/O on non-read/write nodes and potentially crashing the cluster.

• Fixed an issue where the prepared statement path did not support runtime fallback for adaptive ordering, causing ORDER BY LIMIT queries to select an incorrect sort index and become slow queries.

• Fixed an issue where the prepared statement path did not support adaptive ordering runtime fallback, causing ORDER BY LIMIT queries to select an incorrect sort index and become slow queries.

• Fixed a potential server-process crash when remote_exec_cmd was repeatedly executed through prepared statements in multi-master clusters.

• Fixed an issue where JSON_ARRAYAGG could incorrectly propagate a valid NULL state during rematerialization of LATERAL derived tables, causing TempTable to reuse stale JSON BLOB pointers and crash.

• Fixed an issue where JSON_ARRAYAGG could incorrectly propagate a valid NULL state during rematerialization of LATERAL derived tables, causing TempTable to reuse stale JSON BLOB pointers and crash.

• Fixed X-Engine CTable parallel bypass import (direct load) in, handler re- using stale GlobalCtx causing LOAD DATA retry when cluster crash.

• Fixed an issue where the first refresh of an incremental materialized view encountered a last_commit_lsn exception, causing data inconsistency in subsequent incremental refreshes.

• Fixed an issue where stale TABLE cache state persisted after index_merge on partitioned tables, causing secondary index queries to intermittently return empty results.

• Fixed an issue where column statistics could not be persisted or loaded correctly when database or table names contained special characters or tables contained unsupported column types such as JSON.

• Fixed transparent distributed table in with LIMIT single partition UPDATE due to JSON type conversion, unable to through SQL forwarding execution issue.

• Fixed an issue where JPP keypart markers in LATERAL derived-table scenarios could be overwritten by later non-JPP keyparts, causing ordered-index optimization to switch to indexes without valid keyparts and potentially trigger assertions or InnoDB crashes.

• Fixed an issue where JPP keypart markers in LATERAL derived table scenarios could be overwritten by subsequent non-JPP keyparts, causing ordered-index optimization to switch to indexes without valid keyparts and potentially trigger assertions or InnoDB crashes.

• Fixed an issue where a Local RO node in the primary cluster of a multi-master deployment could not register when switching to read-write because the original read-write node had not completed asynchronous offlining in CC.

• Fixed an issue where X-Engine partitioned tables were still assigned an implicit primary key (IPK), causing LOAD DATA bypass import (direct load) to be rejected due to secondary indexes, and resulting in inconsistency with non-partitioned table behavior.

• Fixed an issue where a read-write node holding an internal global lock could cause read-only replay failures.

• Fixed a potential instance crash if a table was deleted while In-Memory Column Index (IMCI) statistics were being collected.

• Fixed stored procedure within PREPARE/EXECUTE dynamic SQL audit log in, scanned rows displaying as entire call process cumulative value, while non- the statement actual scanned rows issue.

• Fixed an issue where audit logs for dynamic SQL executed with PREPARE and EXECUTE in stored procedures showed the cumulative scanned-row count for the entire procedure rather than the actual count for each statement.

• Fixed a use-after-free issue caused by concurrent access to Security_context from PROCESSLIST_VERBOSE and COM_CHANGE_USER.

• Fixed an infinite loop in rowbuf swapping when IMCI shutdown occurred after sort compaction had preallocated but not processed rowid values.

• Fixed column storing table in write not containing NULL value data when marker improper handling, in specific in scenarios where could cause columnar data incorrect issue.

• Fixed an issue where X-Engine parallel LOAD DATA from OSS wildcard file paths did not push down literal prefixes, causing large directories to undergo full paginated scanning and imports to stall at the file enumeration stage. Also ensured that OSS file list errors are properly returned to the upper layer.

• Fixed an issue where X-Engine parallel LOAD DATA did not push down literal prefixes for OSS wildcard paths, causing full paginated scans of large directories and long stalls during file enumeration. OSS file-list errors are now returned to the upper layer.

• Fixed a use-after-free issue caused by concurrent access to Security_context from PROCESSLIST_VERBOSE and COM_CHANGE_USER.

• Fixed incorrect results that could occur in some derived-table queries after column pruning was enabled.

• Fixed an issue where incorrect marker handling during writes without NULL values could cause incorrect columnar data in specific scenarios.

• Fixed a memory leak that could occur when a multi-master cluster shut down with idle-space statistics enabled.

• Fixed an issue where global log merging could stall during multi-threaded parallel DDL operations (such as OPTIMIZE) in multi-master clusters.

• Fixed an issue where using parameter placeholders (?) in ORDER BY/GROUP BY within prepared statements could cause a server crash. The digest is now properly truncated instead of triggering a crash in this scenario.

• Fixed a syntax error caused by a conflict between WITH CHECK OPTION and archived table queries when creating views, and added compatibility for ARCHIVED as a CTE name.

• Fixed a crash caused by repeated pre-processing of window-function views.

• Fixed creating view when WITH CHECK OPTION and archived table query syntax conflict causing syntax error, and compatible with ARCHIVED as CTE name.

• Fixed bitmap-access errors for DATE and TIMESTAMP IS NULL and IS NOT NULL predicates in parallel In-Memory Column Index (IMCI) CTable execution.

• Fixed an issue where InnoDB FTS ngram phrase queries with LIMIT could miss matching results.

• Fixed an issue where queries containing both ROLLUP and window functions could not be executed by IMCI.

• Fixed process crashes that could occur when CTEs with multiple references ran in parallel.

• Fixed IMCI CTable DATE/TIMESTAMP IS NULL/IS NOT NULL in parallel execution in scenarios where bitmap access error.

• Fixed IMCI CTable TIMESTAMP in non- UTC session under ORC trimming and cross- session cache time zone error.

• Fixed an issue where incomplete IMCI split redo writes prevented checkpoint_lsn from advancing.

• Fixed an issue where XA transaction XID containing null bytes could be misidentified as identical to a different transaction during crash recovery, causing recovery failure and preventing the cluster from starting.

• Fixed potential node crashes when subqueries in parallel query execution triggered dynamic range optimization.

• Fixed an issue where incremental materialized view primary-key map entries were not restored on savepoint rollback, preventing NCI find_key failures and stale materialized view rows after repeated update or delete chains for the same primary key.

• Fixed an issue where MDL locks could not be acquired during the DD SCAN phase, leaving columnstore index data unrecoverable after an error.

• Fixed savepoint rollback when incremental materialized view primary key mapping (IVM pkey map) entry not recovery issue, avoiding repeated execution identical primary key UPDATE/DELETE chain after occurring NCI find_key failure and materialized view data stale issue.

• Fixed com_change_user and show processlist/is.processlist query concurrent situation under causing cluster crash.

• Fixed an issue where X-Engine COLUMN large table CHECK TABLE could not exit promptly when a parallel scan was interrupted.

• Fixed an issue where cleaning up cache with multiple versions of TABLE_SHARE during partition MDL operations could crash the cluster.

• Fixed an issue where Hybrid Plan could reuse expired In-Memory Column Index (IMCI) read sequences when shared read views were reused, causing result-visibility anomalies after extra read-view references were closed. Regression tests now cover CSN compatibility, read-view lifecycle, and fallback validation in MPP and Replica scenarios.

• Fixed an issue where extended statistics incorrectly triggered multi-master partition statistics loading and broadcasting in distributed single-table scenarios.

• Fixed an issue where Hybrid Plan could fail to read data from certain read views in extreme scenarios.

Description

• Fixed a potential deadlock between the X-Engine background statistics thread and foreground DDL operations.

• Fixed an issue where In-Memory Column Index (IMCI) improperly handled markers when writing data without NULL values, which could cause columnar data to be incorrect in specific scenarios.

Description

• Added incremental refresh capability for single-table MIN/MAX/STDDEV/VARIANCE and multi-table aggregation queries in incremental materialized views.

Description

• Fixed an issue where secondary index phantom entries could remain after deleting rows by primary key from X-Engine column store tables.

• Fixed an issue when enabling innodb_stats_persistent after, mp_ddl exception issue.

• Fixed an issue where the background statistics collection thread did not trigger memory reclamation under specific scenarios.

Description

• Added Materialized viewBinlog subscription capability, support under cursor data synchronizing consumption materialized view changing more data.

• Incremental materialized view refresh now supports Hybrid Plan, improving execution efficiency for multi-table inner join scenarios where small incremental changes drive large base table joins.

• X-Engine now supports Geo-Distributed Network (GDN), enabling writes in one region and reads from multiple nearby nodes.

• Added X-Engine partitioned table support TRUNCATE PARTITION operation, hybrid partitions table scenarios also Added support for X-Engine partition execution TRUNCATE operation.

• Added support for distributed single-table multi-master routing, cache invalidation, redistribution, and ePQ.

• Extended column statistics support for more accurate cost estimation, with statistics broadcasting capability so that statistics for all partitions are globally visible.

• In-Memory Column Index (IMCI) Added QuickSelect computation sub-, through O(n) fast selecting algorithm high efficiency support percentile query, avoiding full volume sorting overhead.

• In-Memory Column Index (IMCI) storage default enabling partial component dictionary encoding, Added support for partial component dictionary encoding string performing query rewriting and query Optimized.

• Added support for hybrid partitions table (InnoDB/X-Engine/ORC/CSV) Inplace DDL in line creating, deleting, RENAME secondary index, and CHANGE INDEX OPTION, ALTER INDEX COMMENT operation.

• AutoETL support view / pipeline source updating (Inplace-Update), multiple cluster synchronizing authorization, custom ETL parameter and global metadata synchronizing.

• AutoETL now supports inplace updates for views/links, multi-instance synchronization authorization, custom ETL parameters, and global metadata synchronization.

• Built-in model inference capability is included in the cluster specification fee at no additional cost.

• Added Multi-master cluster broadcast table feature, through BROADCAST keyword creating broadcast table, would table data replication to each DN node, eliminating cross- database JOIN network overhead.

• DLM automatic archiving strategy Added support for RANGE COLUMNS + HASHsub-partition table automatic archiving, support archiving target engine including X-Engine row storing, X-Engine column storing, CSV and ORC.

• Orca feature support Redis Stream data structure and related command.

• TTL feature Added support for Binlog, using at under cursor database data synchronizing, ensuring data consistent.

• Added Orca feature for Redis Geospatial Indices data type related command support.

• Local RO now supports IMCI capability. The primary parameter configuration is as follows, with the remaining parameters following the standard Local RO and multi-master IMCI node settings.

  • imci_share_storage_lsn (ON): IMCI node between whether using LSN dimension holding MPP reading consistent.

  • imci_share_storage_data (OFF): Whether IMCI nodes share columnar data.

  • polar_enable_imci (ON): whether using IMCI feature.

  • use_imci_engine (ON): whether using IMCI engine.

  • imci_enable_checkpoint (ON): whether enabling IMCI Checkpoint.

  • polar_enable_imci_with_mm (ON): Multi-master cluster under whether can using IMCI.

  • mm_is_local_replica (ON): Whether the node is a Local RO node in a multi-master cluster.

Description

• Improved asynchronous execution to enhance write performance in high-concurrency scenarios with binlog enabled.

• Optimized In-Memory Column Index (IMCI)full-text index in scenarios where JSON_OVERLAP function query, Optimized can sufficient advantage using index information precision correct estimation base number.

• When controlling total temporary table disk usage, users with SUPER privileges are excluded.

• Optimized Binlog table_map_event in no longer record Functional Index producing Hidden column, avoiding Canal not compatible with issue.

• Temporary table total disk usage control now excludes users with SUPER privilege.

Description

• Fixed an issue where the ANY_VALUE function on string columns could incorrectly return NULL when executing GROUP BY on tables with high NDV.

• Fixed a crash in multi-master partitioned tables caused by write_result being invoked before close_thread_tables during concurrent DML and DDL operations.

• Fixed an issue where extended statistics incorrectly triggered multi-master partition statistics loading and broadcasting in distributed single-table scenarios.

• Fixed an instance crash caused by write_result being called before close_thread_tables during concurrent DML and DDL on multi-master partitioned tables.

• Fixed table converting partition WITHOUT VALIDATION syntax in source single table existing INSTANT ADD/DROP/MODIFY COLUMN when validation failure.

• Fixed an issue where parallel queries with LIMIT/OFFSET could return incorrect results.

• Fixed an issue where the REGEXP function with binary charset in combined window function derived table evaluation could crash.

• Fixed a crash caused by concurrent access to stale page_hash objects during InnoDB buffer pool online resize.

• Fixed a rare crash when adding partitions to tables with GSI while partition MDL locks are enabled.

• Fixed a crash caused by incompatible exec status deserialization formats when multiple versions of parallel query (PQ) coexisted.

• Fixed incorrect execution plan selection for X-Engine large table RANGE queries due to significant index dive estimation deviation.

• Fixed an issue where DROP DATABASE released locks on replica nodes prematurely, allowing user SQL to access deleted files and crash.

• Fixed a crash where condition pushdown join-equal temporary constructs were incorrectly persisted during repeated prepared statement execution.

• Fixed ALTER TABLE re- setting base table ivm_delta_mv_* DD metadata issue. currently only has in ADD ORDER KEY or ADD INDEX DDL during execution, would retaining ivm_delta_mv_* DD metadata, other DDL would re- setting modifying metadata.

• Fixed InnoDB Buffer Pool in line Resize when page_hash old for object was concurrent access causing crash.

• Fixed vector index in table or In-Memory Column Index (IMCI) was deleting after residue after Dump task issue.

• Fixed DLM archiving task in Release in mode archiving failure when Progress state not correct updating as ARCHIVE_ERROR issue.

• Fixed an issue where X-Engine large table RANGE queries with significant Index Dive deviation could select an incorrect execution plan.

• Fixed a deadlock that could occur between multiple threads within B-Tree operations when parallel queries are enabled on the primary node.

• Fixed JOIN eliminating in traversing Item_cond column table existing NULL Item crash-triggering issue.

• Fixed row-column inconsistency in columnar queries caused by IMCI RowBuf multi-column prefetch generating incomplete pack caches under memory pressure.

• Fixed abnormal result ordering when TempTable sort index incorrectly skipped filesort.

• Fixed an issue where UNION views with constant columns in JOIN scenarios returned incorrect results from Derived Condition Pushdown.

• Fixed an issue where ANALYZE TABLE on full-columnar partitioned tables missed statistics collection for pruned partitions.

• Fixed an issue where TempTable sort indexes could incorrectly skip filesort, causing abnormal result ordering.

• Fixed an issue where dbms_oss.delete_table_file did not invalidate the OSS global table list cache after deleting a partition or table file, causing dbms_oss.list_table() to still display the deleted partition.

• Fixed an issue where queries using UNION or UNION ALL with GROUP BY could fail when subqueries returned constant strings.

• Fixed a potential deadlock between multiple threads in B-Tree when parallel query was enabled on the primary node.

• Fixed an issue where shared read views in Hybrid Plan scenarios could use expired IMCI read sequences, causing result visibility anomalies.

• Fixed an issue where repeated invocation of stored procedures could cause a process crash due to incorrect reuse of SQL Hint (such as SET_VAR) related memory.

• Fixed an issue where IMCI statistics persistence timestamp comparison failed across multiple columnar nodes.

• Fixed an issue where after fast conversion of a regular table to a partitioned table (WITHOUT VALIDATION), only the first partition retained the In-Memory Column Index (IMCI) while the remaining partitions lost it.

• Fixed an issue where Window Functions could not use IMCI for execution in certain scenarios, and GROUP BY pushdown could prevent queries from being executed in IMCI.

• Fixed an issue where statistics collection missed pruned partitions when executing ANALYZE TABLE on fully columnar partitioned tables.

• Fixed Online DDL failure when tables contain virtual column indexes.

• Fixed a rare instance exception caused by MDL release timing issues between background worker and foreground leader threads during DLM automatic archiving.

• Fixed an issue where the DLM archiving task progress status was not correctly updated to ARCHIVE_ERROR upon archiving failure in Release mode.

• Fixed a potential memory leak in LRU_RB_BLOCK during IMCI RowBuf cache insertion failure.

• Fixed an issue where converting a regular table to a partitioned table (WITHOUT VALIDATION) with columnar indexes resulted in only the first partition retaining the columnar index while others lost it.

• Fixed a potential primary node lockup during X-Engine columnar table writes.

• Fixed validation failure in table-to-partition conversion with WITHOUT VALIDATION when the source table has INSTANT ADD/DROP/MODIFY COLUMN history.

• Fixed a crash when executing ST_Buffer on a POINT with an extremely small distance in geographic coordinate systems (SRID 4326).

• Fixed a crash caused by dirty writes in IMCI RowBuf metadata compaction due to incorrect handling of page size overflow paths.

• Fixed an instance exception when executing DLM automatic archiving on partitioned tables with virtual columns.

• Fixed an issue where Geographic sitting standard system (SRID 4326) under for POINT execution ST_Buffer transmitting entering extreme small distance when could cause a crash.

Description

• Added built-in model free inference capability, support model in in-database direct inference, no token consumption, no additional charges.

Description

• Fixed an issue where, for INTERVAL partitioned tables, specifying ROW_FORMAT=COMPRESSED without explicitly setting KEY_BLOCK_SIZE caused the ROW_FORMAT attribute to be lost.

Description

• Added materialized view Parallel refresh capability.

• Added Incremental materialized views feature, in full volume materialized view base foundation on evolution advancing, improving refresh performance.

• The new database and table restore workflow enables fast full data queries that complete in minutes, regardless of the amount of data being restored.

• Added support for asynchronous cleaning up DROP/TRUNCATE PARTITION residue expired Global Secondary Indexes (GSIs) data feature.

• Added GDN multiple writing edition, providing table level alias full global multiple point write capability.

• Added ask_polar expression support column storing plan capability.

• Added an independent concurrency control module for vector search queries, decoupled from In-Memory Column Index (IMCI) query concurrency control.

• Upgraded vsag to the official 0.16 branch version. Added parallel structure building for the HNSW algorithm and support for creating indexes based on the hgraph algorithm.

• Added support for IVF and IVF PQ vector index types in vector search.

Description

• Optimized What is In-Memory Column Index (IMCI) Hash Join performance, through would Join in between results converting ization as row format storage, reducing data materialized with to performance overhead.

• Optimized the X-Engine storage engine by removing unnecessary implicit primary keys (Implicit PKs) to simplify the data update logic and in-memory data structures in the storage layer.

• Optimized the table cache invalidation strategy during ADD/DROP PARTITION operations on partitioned tables to reduce the impact on queries.

• Optimized the concurrent computation logic for vector index builds, allowing more threads to be used to accelerate index creation.

• Fixed inaccurate adaptive rate limiting for DDL bulk load and improved the adaptive rate-limiting algorithm.

Description

• Fixed a crash that could occur during an incremental refresh of a materialized view on a hybrid partitioned table when reading partition metadata.

• Fixed an issue where ONLINE DDL failed when the table contained a virtual column index.

• Fixed an issue in the community edition of MySQL where purging binary logs at startup could result in an incorrect Previous_gtid for the global transaction identifier (GTID).

• Fixed an issue where INSERT ON DUPLICATE KEY UPDATE could update the wrong row during concurrent inserts from multiple sessions into a table with an auto-increment column.

• Fixed an issue where a columnar query using Hybrid Index Search with a LIMIT clause could fall back to row-store execution if the LIMIT OFFSET value was too large.

• Fixed various incompatibility issues on multi-primary nodes when Parallel Parse is enabled.

• Fixed a potential memory leak in VSAG-based HNSW and DiskANN indexes within the vector index of a columnar index (IMCI).

• Fixed What is In-Memory Column Index (IMCI) mixed combined execution plan (Hybrid Plan) could cause reading view (Read View) reclaiming when cluster crash.

• Fixed a memory leak of approximately 626 KB that was caused by the buffer pool not being released in the initialization abort path when an InnoDB HA failover failed to start.

• Fixed a memory out-of-bounds access issue caused by concurrent access between rollback segment (Rsegs) lock objects and the Redo log replay thread during InnoDB HA failovers, and a memory leak in the error branch for large records (big_rec) in the clustered index insertion path.

• Fixed a memory leak in the sql_sharing and sql_sharing_proc modules of SQL Plan Management (SPM) where internal buffers of string objects were not properly released.

• Fixed a use-after-free issue in the X-Engine storage engine that could occur if a background space reclamation task ran concurrently with a database instance shutdown.

• Fixed an invalid memory free operation and an indirect memory leak in InnoDB caused by the LatchCounter incorrectly releasing an embedded counter object upon destruction.

• Fixed a minor memory leak that occurred with every request in the ask_polar module because the request header linked list was not properly released after an HTTP request completed.

• Fixed a use-after-free issue where the performance monitoring module (performance_schema) could access a stale pointer to a session thread (THD) during a concurrent scan after the thread had already released its resources.

• Fixed multiple memory leaks in the extended statistics (polar_extend_stats) module. The issues were caused by improperly destroyed sampled cache data, a value-tuple buffer detached from memory management, and missing ownership in a filter-effect clone container.

• Fixed a memory out-of-bounds access issue during instance shutdown. The issue occurred when a redo log synchronization thread accessed a freed resource because a synchronization object was destroyed during binlog cleanup before the thread had completely exited.

• Fixed a stack overflow in the X-Engine storage engine. The issue occurred when a DDL operation on a partitioned table generated a full table name that exceeded the system's maximum length and was written to a fixed-size buffer.

• Fixed an instance crash that occurred during a parallel refresh of a full refresh for materialized view in specific semijoin query scenarios.

• Fixed a memory out-of-bounds access issue that occurred when a Prepared Statement was recompiled, and a table name pointer in a query hint still pointed to a freed memory area.

• Fixed an issue when executing LOAD DATA statement due to toward NOT NULL column import NULL value while error after, when before session field validation state not was correct re- setting, causing same one session in subsequent valid INSERT statement also mistake reporting Column cannot be null error. re- new building establishing connection can temporary when bypassing.

• Fixed an issue where prepared statements (Prepared Statement) during execution transmitting entering non- method day period when between value, in generating warning information when occurring out-of-bounds write, triggering buffer overflow and cluster crash.

• Fixed an issue where parallel queries could cause a system crash when ORDER BY referenced a constant alias expression.

• Fixed a crash caused by incomplete copying of an internal data structure when the query optimizer pulled up a subquery (merged it into the outer query). Also fixed an issue where the table cache pointer was null or invalid.

• Fixed an issue where a columnar index became invalid and unusable after a DDL operation that simultaneously added both a columnar index and a regular secondary index to a partitioned table.

• Fixed single table fast converting partitioned table and other PolarDB special has ALTER syntax and DMS comment compatible with issue, thereby resolving PolarDB to PolarDB DTS transmitting transport issue.

• Fixed a dangling pointer access issue caused by the PlanHintsStatement not being reset when a Prepared Statement was recompiled.

• Fixed an incorrect string escaping issue that could occur when PolarDB AutoETL created a search view containing an ETL_UDF.

• Fixed a memory issue that could occur on a read-only node (RO) when flushing the cache if the extended statistics feature was enabled.

• Fixed cluster version upgrade process in, system table mysql.ttl_job_history scan_cost, purge_cost, purge_rows field type not compatible with causing data write failure.

• Fixed an issue where the table row lock wait time for the X-Engine storage engine was not included in the lock wait statistics of the slow query log.

• Fixed an issue where the log allocator (log_allocater) was not released in time on Standby HA nodes running as Replica, which could cause memory anomalies.

• Fixed an issue where threshold calculation was incorrect when processing multi-column (Row Expr) IN list statements converted to semijoin.

• Fixed an issue where the query cache triggered InnoDB strong consistent read waits prematurely, causing In-Memory Column Index (IMCI) read-only node replication delay timeouts.

• Fixed a data anomaly caused by a lost Header Page Apply when Transparent Data Encryption (TDE) was used on a read-only node (RO) or a standby node.

• Fixed an issue where the Thread Pool and Connection Control plugins were started simultaneously on new nodes due to incorrect parameter configuration. A conflict check will now be automatically performed at startup.

• Fixed a buffer overflow and instance crash that could occur when querying a partitioned table using a secondary index on an integer field, if the first InnoDB partition of the table was a Partition Partial Index.

• Fixed a crash in the community edition of MySQL caused by an exception when updating a spatial index that is a secondary index.

• Fixed Orca feature DUMP command results not was low version (<7.4.0) Redis identifying alias issue.

• Fixed What is In-Memory Column Index (IMCI) vector index after dump task in table deleting or index deleting after unable to and when final stopping issue.

Description

• Added built-in model free inference capability, support model in in-database direct inference, no token consumption, no additional charges.

Description

• Added mm_enable_wait_binlog_flush parameter, support What is Multi-Master Cluster (Limitless) and other pending Binlog disk flush completing, to resolving Binlog asynchronous disk flush could cause data consistent issue.

• Fixed What is Multi-Master Cluster (Limitless) using cold data archiving feature as CSV after tables, execution DROP TABLE when cluster crash.

Description

• Fixed an issue where a minor version upgrade to 8.0.2.2.34 would get stuck during the upgrade task flow due to a problem in the upgrade process.

Description

• Added multiple index statistics sampling strategies that can be enabled via parameters.

• Added support for the REBUILD PARTITION operation on partitioned tables.

• What is In-Memory Column Index (IMCI) Hybrid Plan (including Hybrid Index Search and Hybrid Index Join) Added support for MPP execution mode support.

• Added support for DiskANN index type in vector search.

• Added the parallel Copy DDL feature, which significantly improves the execution efficiency of COPY DDL operations.

• Added PolarSearch searching testing graph (Search View) feature, implementing PolarDB for MySQL to PolarSearch Auto-ETL capability, improving data synchronizing convenient fast.

• Added the WITHOUT VALIDATION option for REORGANIZE PARTITION, supporting fast modification of RANGE partition boundaries. This is suitable for scenarios where partition data is approaching maximum capacity, enabling quick boundary changes to route new data to new partitions.

  Note

  Limits: single DDL only support modifying one partition edge boundary value, and not supported more modifying partition name.

• Added a parameter xengine_default_storage_table_format, support specifying creating What Is X-Engine table default table format.

• Added support for Full refresh of materialized views refresh when would query unloading to specifying node on execution.

• Traditional InnoDB compressed table format has a data corruption risk when BLOCK_SIZE is less than 4. Added the polar_compressed_table_policy system variable to control the compression policy for newly created tables.

  • NATIVE: source sample execution.

  • ADJUST_KEY_BLOCK_SIZE (default): Allows table creation and automatically corrects the KEY_BLOCK_SIZE if it is less than 4.

  • AUTO_CONVERT_ROW_FORMAT: Silently converts the table to the default row format.

  • REJECTED: Rejects the operation and returns an error.

  Note

  • This policy applies only to newly created compressed tables and does not affect existing compressed tables.

  • For clusters that run a kernel version earlier than 8.0.2.2.34, we recommend that you no longer create InnoDB compressed tables with a BLOCK_SIZE less than 4.

• Added support for hybrid partitioned tables with primary keys, allowing partitions to be archived to OSS partitions.

• Added a DEMAND mode for SQL Plan Management (SPM) that can be used to inject, capture, and manage execution plans.

• Added support for PolarDB three-layer decoupling (CXL) memory pool dynamic enabling and disabling. currently only support entire CXL memory dynamic switch, temporary not supported CXL memory pool size dynamic changing ization.

• Added support through ask_polar expression directly connection A inside Cloud large model service flat Bailian capability.

• Added support for multiple passes within a statement. Also fixed a crash issue with adaptive ordering in specific scenarios.

• Added LNNVL expression, suitable configuring row column storing execution.

Description

• Optimized partition archiving to OSS partitioned table (cold data archiving) during the process of, UPDATE/DELETE statement automatic pruning currently in archiving partition and OSS cold data partition, most large limit degree avoiding DML statement was blocking.

• Optimized What Is X-Engine in restart phase recovery metadata data structure speed, reducing restart consumption when.

Description

• Fixed an issue when enabling Query Cache in cases where Full refresh of materialized views refresh after cannot and when reading to most new results issue.

• Fixed partial in scenarios where ORDER BY sound clear sequence order not take effect issue, meanwhile Optimized grouping query in Loose Index Scan (LIS) called word processing.

• Fixed an issue when enabling thread_pool_enabled parameter after, performance_schema memory statistics incorrect issue.

• Fixed an issue where occasional transaction loss for In-Memory Column Index (IMCI) under the transaction persistence feature of Seamless Switch led to data inconsistency between row and column stores.

• Fixed ORC format timestamp type min/max value single position not for aligned and ignoring accepting seconds partial issue.

• Fixed an issue where the default value became invalid due to a memory exception when an instant add column operation that adds a column with a long string default value was converted to an asynchronous operation.

• Fixed JPPD feature for Item_ref for object cloning (clone) implementing not improving causing subsequent crash.

• Fixed an issue in columnar index (IMCI) where a filter condition was incorrectly pushed down past a LIMIT clause, which could lead to incorrect results.

• Fixed mysqlredolog working having parsing specific MLOG_METADATA_CHANGE record when crash.

• Fixed an issue where JSON_TABLE could probabilistically crash in the trigger execution path.

• Fixed an error that could prevent statistics from being updated when collecting statistics for a columnar index (IMCI) with the low-cardinality column sampling enhancement (imci_stats_scan_max_ndv) enabled.

• Fixed concurrency control (CCL) for DELETE xx FROM ... syntax cannot take effect issue (not affecting DELETE FROM ... syntax).

• Fixed an incorrect memory state on a read-only (RO) node when no extensions were initialized.

• Fixed an issue where the query filter did not take effect when using a Bloom Filter with a date type specified.

• Fixed What is Orca (Redis-Compatible) cleaning up expired KEY using pcur could cause a crash, replacing as storage KEY method.

• Fixed an issue where an int integer overflow in the In-Memory Column Index (IMCI) NCI file system after long-running operations caused out-of-bounds memory access and resulted in IMCI transaction replay getting stuck.

Description

• Fixed Online Copy DDL in certain in scenarios where not take effect issue, meanwhile adding in What is Multi-Master Cluster (Limitless) and other in scenarios where limiting.

• Fixed an issue where inconsistent internal purge sequence synchronization between the read-write (RW) node and read-only (RO) node when using X-Engine columnar tables could cause intermittent crashes on read-only nodes.

• Fixed a slow query issue that occurred when a columnar index (IMCI) created empty statistics for an OSS partitioned table (cold data archiving) because it calculated the actual row count.

Description

• Added a native BSON data type. This type supports projection, update, query, and aggregation operations.

• Added the ability to call a large language model (LLM) directly by using standard SQL statements.

• Added Full refresh of materialized views refresh process in parallel inserting capability. through parallel processing, significantly improving materialized view data refresh speed.

• Added the ETL feature, which can automatically perform data extraction, transformation, and loading to improve data synchronization and aggregation efficiency.

• Added configurable execution time windows for SQL-layer TTL. You can now specify start and end times for TTL tasks to automatically clean up expired data within a preset time frame.

• Added What Is X-Engine for column storing storage format support, support low becoming local storage and high performance analysis.

• Added support for executing DDL operations in COPY mode on hybrid partitioned tables that contain OSS partitions.

• Added What Is X-Engine support index column character set and collation, utf8_unicode_ci and utf8mb4_unicode_ci.

• Added Full refresh of materialized views full volume refresh node minute flow feature. support base at cost would refresh task minute sending to specific node execution, to interval leaving resource affecting.

• Added hybrid partitions table support What is In-Memory Column Index (IMCI) feature. current in, InnoDB+X-Engine, InnoDB+ORC, InnoDB+CSV group combined hybrid partitions table can through ALTER TABLE COMMENT statement management In-Memory Column Index (IMCI), thereby advantage using IMCI accelerating for partitioned table analysis query.

• Added Overview feature for DATETIME data type support. removing TIMESTAMP external, current in also can using DATETIME type column to defining data automatic expired strategy, to satisfying more broad business scenarios.

• Added What is Orca (Redis-Compatible) compatible with under for LUA footnote local support. now supports EVAL, EVALRO, SCRIPTS and other related command.

• Added vector index incremental structure building feature. the feature allowing index following data write while actual when updating, significantly improving sea volume toward volume data entering database and updating efficiency.

• Added Eliminate redundant GROUP BY/ORDER BY feature, through for GROUP BY and ORDER BY clause performing ization simple and eliminating, reducing redundant remainder computation, improving specific SQL query execution efficiency.

Description

• Optimized physical replication efficiency by enabling parallel Redo log parsing on read-only nodes to reduce primary-replica lag.

• Optimized What is In-Memory Column Index (IMCI) restart when between. modifying creating NCI (Non-Clustered Index) module causing its support Checkpoint mechanism, thereby eliminating can eliminating LOADING NCI full volume structure building NCI step step.

• Optimized What Is X-Engine metadata data management mechanism. through cycle period releasing manifest log and old version manifest checkpoint, avoiding due to metadata data file holding continued increasing long while causing quota external storage empty between occupying using.

• Optimized the performance of partitioned vector retrieval for columnar index (IMCI). This optimization reuses distances calculated during the vector index search phase in the multi-way merge to avoid redundant calculations. Also fixed an issue where a condition pushdown for a columnar index (IMCI) could return incorrect results in ORDER BY LIMIT scenarios.

• Optimized What is In-Memory Column Index (IMCI) pointer for query condition in containing exceeding long string type IN-List query performance.

• Optimized the query plan search process to accelerate the search for the join order of multiple tables, which reduces the overall optimization time for complex SQL queries.

• Optimized grouping aggregation combined (HashGroupby) execution efficiency. through introducing toward volume ization probing checking hash table operation, in improving grouping aggregation combined performance meanwhile, slightly reducing computation process in memory consumption.

• Removed the limit on the number of partitions for a table.

• Optimized What is In-Memory Column Index (IMCI) query execution strategy, default enabling adaptive should grouping aggregation combined execution. the feature can based on data minute distributing and query feature automatic selecting most high efficiency aggregation combined algorithm, thereby improving containing GROUP BY clause analysis query performance.

• Optimized What is In-Memory Column Index (IMCI) in Hash Join computation sub- execution efficiency. this time Optimized for all using Hash Join query all has performance modifying good, especially its for Semi Join scenarios performance improving significantly.

• Optimized the performance of the distributed memory pool by adding relevant parameters and monitoring items.

Description

• Fixed a crash that could occur in specific scenarios when a filter condition containing a trigger field was pushed down to a derived table.

• Fixed an issue where condition pushdown optimization was ineffective because a variable assignment function was present in the query.

• Fixed Global Secondary Indexes (GSIs) in specific LIMIT OFFSET under pushing Optimized in scenarios where could cause a cluster crash.

• Fixed a stability issue that could be triggered by improper internal metadata handling during the upgrade process for certain partitioned tables.

• Fixed an issue where executing DML operations on a hybrid partitioned table that contains OSS partitions could block concurrent reads and writes.

• Fixed an issue where the high-performance HashJoin operator processing RIGHT JOIN with data spilling to disk could cause loss of result set rows.

• Fixed a rare issue where replication could get stuck in a columnar index (IMCI) in scenarios with frequent DDL operations, leading to data latency on the node.

• Fixed an issue where statistics could not be collected as expected for string columns with wide character sets (such as utf8), even if the defined length was short.

• Fixed an issue where What is Multi-Master Cluster (Limitless) in, execution Global Secondary Indexes (GSIs) related DDL during operations, due to and after thread conflict while could cause a cluster crash.

• Fixed an issue in the What is Orca (Redis-Compatible) mode where data table metadata might not be updated in a timely manner.

• Fixed a potential SQL parsing error that could occur when using dbms_outln.add_optimizer_outline_sharding to create a Statement Outline for sharded tables.

• Fixed an issue where PCUs for a low-specification, no-load serverless node with columnar index (IMCI) would repeatedly scale up and down.

• Fixed a potential out-of-memory (OOM) issue in the serverless feature for columnar index (IMCI) that could occur when resources were rapidly scaled up and then scaled back down, because the cache quota was not reclaimed in a timely manner.

• Fixed large amount of frequently volume table in What is elastic parallel query estimation cost error, advancing while causing cluster crash.

• Fixed a crash that could occur when defining a partition range for an Interval partitioned table if a time type was specified.

• Fixed What is In-Memory Column Index (IMCI) unable to execution query condition in containing system changing volume SQL statement issue.

• Fixed an issue where statistics collection could fail on tables with frequent updates and deletions, and also slightly improved sampling efficiency.

• Fixed an issue where memory usage could continuously increase if a DDL operation was performed during incremental log application.

• Fixed an issue where a table could not be opened for an Index Merge plan when the query did not need to access the OSS partitions of a hybrid partitioned table.

• Fixed a stability issue that could cause a crash when the buffer pool was much smaller than the intermediate result set required by a query.

Description

Fixed What is Orca (Redis-Compatible) restore command in recovery already expired hash field when would legacy missing its TTL (generating storing when between) information issue.

Description

• Added support for Join condition pushdown to containing UNION clause derived table (derived table).

• Added cost-based OR expansion query rewrite, which can intelligently convert OR or IN expressions to UNION ALL for better index utilization and improved performance in specific query scenarios.

• Added support for vector index in line DDL support, allowing in not modifying changing In-Memory Column Index (IMCI) state in cases where, new building, deleting, modifying and re- building already has vector index.

• Added Configure transparent data encryption (TDE) support SM4 algorithm, as data providing more high and other level adding dense maintaining protecting.

• Added global only reading column storing node support Multi-master sharded tables.

• Added support for vector search adaptive should execution Optimized, support dynamic selecting execution method (prefilter/postfilter/inline filter).

• Added snapshot update feature, through introducing compatible with PostgreSQL updating mechanism, Optimized specific in scenarios where UPDATE operation execution efficiency. the feature through parameter innodb_polar_use_snapshot_update controlling, support session level and global level enabling.

• Added What is In-Memory Column Index (IMCI) support full-text index, valid improving text local fuzzy searching and analysis performance.

• Added automatic archiving cold data (DLM) Added support for with has What is In-Memory Column Index (IMCI) partitioned table archiving to OSS external table, implementing DLM and IMCI compatible with.

• Added monitoring metrics for Orca (Redis-compatible) to enhance its observability, including active connections, new connections, memory usage, and the number of pending tasks.

Description

• Optimized the Page Apply efficiency in physical replication. This reduces I/O overhead when read-only nodes access cold data, minimizing sporadic latency jitter and enhancing cluster stability.

• Optimized the predicate deduction (deduce_predicates) logic in the query optimizer to more effectively identify static conditions on partition keys, enabling more aggressive partition pruning and improving query efficiency.

Description

• Fixed an issue where executing RENAME TABLE and other DDL operations on tables containing views after a high-availability failover could cause a Duplicate entry error.

• Fixed an issue where the database table recovery feature could not correctly recover a table that contained columns deleted using the instant drop column method.

• Fixed an issue in a multi-primary cluster where deleting a read-write node was interrupted after an access point switchover.

• Fixed What Is X-Engine partitioned table execution adding or deleting partition during operations, need adding table level MDL lock issue.

• Fixed What is Orca (Redis-Compatible) in deleting key (key) after establishing i.e. write, could cause dbsize statistics not accurate issue.

• Fixed an issue where dynamically disabling the thread pool parameter loose_thread_pool_enabled could cause connections to hang after DDL execution completed.

• Fixed What is In-Memory Column Index (IMCI) in using ETL when using the feature not supported JSON TABLE issue, and Optimized IMCI query memory not sufficient when error submitting showing information.

• Fixed an issue where a SELECT statement containing both IN-List and aggregate functions with DISTINCT could prevent the query from being executed on In-Memory Column Index (IMCI).

• Fixed an issue in IMCI that could cause inaccurate multiplication results when processing DECIMAL data with more than 16 digits after the decimal point.

• Fixed an issue with the IMCI Serverless feature where proactive scale-out was continuously triggered due to an abnormal index status.

• Fixed an issue where a standby node could hang during the recovery phase in Point-in-Time Recovery (PITR) and database and table recovery scenarios.

• Fixed an issue that could cause existing Binlog dump connections to hang and affect Binlog cleanup after a read-only node was promoted to the primary node.

• Fixed an issue in read/write (RW) node on execution SHOW POLAR STANDBYS when executing the command, query to standby using (Standby) node latency seconds number (delay_second) not accurate issue.

• Fixed an issue where the query optimizer's column pruning feature could cause the process to crash by incorrectly removing a subquery that was referenced by multiple projection columns.

• Fixed an issue where inaccurate statistics for non-BIGINT integer columns prevented the query optimizer from generating an optimal execution plan.

• Fixed an issue where querying JSON data of a specific length could return the "The JSON binary value contains invalid data" error on a read-only node when the innodb_blob_slave_read_consistent parameter is enabled.

• Fixed an issue that could prevent queries containing a recursive Common Table Expression (CTE) from running on an IMCI.

• Fixed an issue where a newly registered read-only node could report an SDI modification failure when executing DDL operations after being promoted to the primary node.

Description

Added dynamic switch, support dynamic enabling only reading (RO) node cursor standard position setting adjusting feature.

Description

Fixed an issue where HA (high-availability failover) during uninitialized expansion under high-concurrent write pressure could cause the cluster to crash.

Description

Fixed an issue where In-Memory Column Index (IMCI) Hybrid Plan wide table query acceleration unexpectedly fell back to row store execution due to query optimization (converting AVG functions to SUM/COUNT).

Description

• Added the ability to configure whether the display width attribute is shown for integer fields (BIGINT, INT, MEDIUMINT, SMALLINT, TINYINT, etc.) when the ZEROFILL attribute is not defined.

• Added concurrency management and status query features for background vector index creation, enabling effective resource consumption control and real-time progress tracking when creating large-scale vector indexes.

• Added CXL memory extending feature, allowing through CXL (Compute Express Link) memory pool as cluster dynamic adding buffer pool capacity volume, to improving large scale data processing performance and system throughput.

• as What is In-Memory Column Index (IMCI) adding base at FAISS database HNSW and HNSWPQ two kind vector index type, to support more high performance, more low memory usage high dimension toward volume near similar most near adjacent (ANN) retrieval.

• in What is In-Memory Column Index (IMCI) vector search in Added inline filter mode, implementing standard volume filtering and vector search single computation sub- integrating combined computation.

• Added support for Oracle-compatible synonyms (SYNONYM). You can now create synonyms for tables or views and operate on them directly.

• Extended the sub-partitioning feature to support RANGE COLUMNS, LIST COLUMNS, and LIST COLUMNS+DEFAULT partition types, with support for DATE, DATETIME, CHAR, and other sub-partition key fields. All combinations of primary and secondary partition types are now supported, with partition pruning optimization for related queries.

• Added support through What is In-Memory Column Index (IMCI) accelerating refresh full volume materialized view, as re- miscellaneous query providing pre- computation results aggregation, significantly improving high frequency aggregation combined or associated query response speed.

• Added the embedding expression to convert input text to vectors in real time. Also supports automatic conversion of text content to vectors during data writes through materialized virtual columns.

• as INSERT ... VALUES/SET ... ON DUPLICATE KEY UPDATE syntax adding alias name (alias) support, allowing in UPDATE clause in through alias name referencing using INSERT partial Added data, causing SQL logical more clearing clear, simple clean.

• Extended the cold data archiving feature to support archiving cold data partitions in partitioned tables to the X-Engine storage engine. After archiving, the table becomes a hybrid partitioned table of InnoDB and X-Engine, supporting unified queries across different storage engine partitions.

• Added support for non- What is In-Memory Column Index (IMCI) node on non- RU (Read Uncommitted) interval leaving level alias session strong controlling forwarding feature. through enabling imci_ignore_unsupported_isolation_level parameter, allowing this some session request was forwarding to column storing node execution, to advantage using column storing accelerating capability.

• Added json_arrayagg expression, using at in What is In-Memory Column Index (IMCI) query in would multiple row results aggregation combined as one JSON number group, improving for JSON format data processing capability.

• Integrated Resource Control with the Orca feature (Redis-compatible). This allows you to manage the overall CPU resource utilization of the Orca system by using a parameter, which defaults to 100%.

• as Orca feature (compatible with Redis protocol) Added support for PUBSUB command support, including PUBLISH, SUBSCRIBE, UNSUBSCRIBE, PSUBSCRIBE, PUNSUBSCRIBE, PUBSUB CHANNELS, PUBSUB NUMPAT and PUBSUB NUMSUB, extending eliminating information sending distributing and subscription capability.

• The slow query log feature now supports sampled recording of actual SQL statements and storing them in shared SQL, enabling database administrators to precisely troubleshoot and optimize excessively long or high-frequency slow queries.

• Added a parameter rds_max_tmp_disk_space, using at controlling large query producing temporary table for local disk empty between total using volume, preventing disk was occupying full.

• Added in table structure deleting after, dbms_oss.delete_table_file support cleaning up OSS on partitioned table for should file.

• Added Time to live (TTL) feature current already COPY category DDL syntax. currently, only support INPLACE category DDL.

Description

• Optimized Row and Column Routing Guide cost estimation model, allowing its base at string row execution overhead performing computation, while non- depending on parallel Optimized after cost, thereby in specific in scenarios where do outputting more optimization execution plan selecting.

• Added a feature to measure the time consumed by internal operations of row-store and column-store execution operators. This data supports coefficient fitting and automatic optimization of the row-column cost model.

• as Statement Outline Added record source initial SQL statement feature, Added support for when before loop environment in based on source initial SQL re- new generating Outline, improving execution plan management and fixed ization flexible active.

• Restructured the command processing structure and optimized lock paths, improving overall read and write performance by approximately 10%.

• Optimized partitioned table query performance. Fixed an issue where new partition key constant conditions generated during the query optimization phase (such as MEP) could not trigger secondary partition pruning, and enabled partition pruning for inner tables in nested outer joins.

• Optimized the processing logic for union queries with a large number of query blocks, resolving an issue where the find_cloned_select_lex function consumed excessive time and fully loaded a single CPU core.

• Optimized the metadata loading process when a column store node restarts in large-scale data scenarios. This prevents out-of-memory (OOM) issues caused by an overly long metadata list.

• Unified the local file naming convention for related components so that file names no longer vary by node role, simplifying file management during high-availability failover.

• Optimized the logic for the background elastic thread pool of a read-only column store node when it obtains task status, reducing its CPU resource consumption.

Description

• Fixed an issue where a parallel query could cause a cluster to crash in specific condition pushdown scenarios.

• Fixed an issue where data inconsistency could occur on a read-only node in some scenarios when reading large JSON fields or frequently partially updated JSON fields.

• Fixed an issue where NCI (Non-Clustered Index) failed to automatically trigger capacity expansion when insertion into the LRU cache failed.

• Fixed abnormal storage space growth caused by vector index data files not being cleaned up promptly.

• Fixed a memory error caused by abnormal initialization of the dec_buffs member in the Item_sum_sum::resolve_type function during the row store execution phase of aggregate function queries.

• Fixed issues in IMCI where vector retrieval queries could incorrectly fall back to row-store execution and where the query optimizer made poor cost-based choices for scalar filter queries.

• Fixed an issue with listing OSS files using the dbms_oss.list_table_file command.

• Fixed an out-of-memory (OOM) issue on read/write (RW) nodes when performing an ONLINE DDL operation on a table with an IMCI. This was caused by the default batch write behavior.

• Fixed an issue where DDL operations could cause a cluster crash when the ORC hybrid partitioned table name contained special characters.

• Fixed an issue where a read-only node might not find some records during B-tree index structure adjustments under high write pressure.

• Fixed an issue where adding an In-Memory Column Index (IMCI) to a table without a primary key but containing a UNIQUE KEY could cause cluster exceptions.

• Fixed an issue when dbms_dlm.archive_partition2table archiving partitioned table data to OSS external table when permission validation issue, current already reducing permission need to requesting, only requiring INSERT, CREATE, DROP, ALTER permission i.e. can operation.

• Fixed What is Orca (Redis-Compatible) in, LPOS command in MAXLEN option value exceeding column table local body length when causing cluster crash.

• Fixed a memory leak that could occur when reading data from a cold storage table in ORC format.

• Fixed partial query due to memory using volume computation not accurate, causing partial query returns OOM Error while execution failure.

• Fixed an issue where query performance could degrade in Multi-master Cluster (Limitless) Edition under high pressure.

• Fixed an issue where constant condition propagation failed during query optimization due to inconsistent character set lengths.

• Fixed What is Multi-Master Cluster (Limitless) in aggregation combined node on execution CREATE ... IF NOT EXISTS statement after, gtid_executed occurring empty insight issue.

• Fixed What is Multi-Master Cluster (Limitless) primary standby when nodes high-availability failover period between, could cause a cluster crash.

• Fixed an issue where an IMCI query could return incorrect results if it contained both the DISTINCT and AVG aggregate functions.

• Fixed an issue caused by Item_func_conv_charset result_field in SQL ending after not cleaning up, causing re- using when cluster crash.

• Fixed What is Multi-Master Cluster (Limitless) under, local only reading (RO) node hot failover as reading writing (RW) after nodes, for What is In-Memory Column Index (IMCI) table execution DML operation could trigger RW node crash.

• Fixed an issue caused by dangling pointer access when an index was released during an INSTANT DDL operation on an IMCI.

• Fixed parallel query in multiple layer derived table nesting scenarios in, could trigger cluster crash.

• Fixed an issue where Data Cleanup (TTL) information was lost during a CREATE PARTITION TABLE operation because the main Data Dictionary (DD) information was not updated.

• Fixed Adaptive execution plan switching (adaptive ordering) Optimized in LIMIT OFFSET statement on unable to correct adaptive should issue.

• Fixed a sorting error that could occur when using ORDER BY DISTANCE for vector similarity sorting with distance values close to 0.

• Fixed an issue in Elastic Parallel Query where cross-node MPP execution could cause incorrect results or a cluster crash when multi-level nested shared subqueries were present.

• Fixed an issue where column pruning optimization could cause a cluster to crash when a query contained a recursive Common Table Expression (CTE).

• Fixed an issue where the Query Cache could access invalid memory under multi-threaded concurrent access.

• Fixed an issue where setting expiration time for Hash fields with duplicate fields could cause a cluster crash.

• Fixed What is Multi-Master Cluster (Limitless) in high-availability failover after, could cause cluster unable to starting or data unable to holding long ization issue.

• Fixed an issue where duplicate data could be retrieved when accessing a spatial index on a read-only node.

• Fixed an issue where the execution results of the IMCI json_valid function could be inconsistent between the row store and column store when the first parameter is of the varchar type.

• Fixed an issue where duplicate data appeared in the result set after an EXISTS subquery was converted to a SEMI JOIN (with the duplicate weedout strategy) when parallel query was enabled.

• Fixed an issue in the IMCI shared storage architecture where the column index of a cold storage table on a read-only column store node could be missing.

• Fixed an issue where a cluster could occasionally crash after adaptive aggregation was enabled in queries from older versions.

• Removed unnecessary checks for compressed row status in the code to avoid triggering unexpected errors in special cases.

• Fixed an issue where a read-only column store node could experience replication latency in scenarios with frequent DDL operations.

Description

• Fixed an issue in the In-Memory Column Index (IMCI) feature where a node could crash when the deltastore executes a swap task, due to the pax page length overflowing uint32 or encountering a null pointer.

• Fixed an issue where improper global variable initialization order caused In-Memory Column Index (IMCI) one-write-multiple-read Follower file operations to be unexpectedly blocked.

• Fixed a rare bug that could cause the background purge process to corrupt a secondary index during a transaction rollback.

Description

• Optimized What is In-Memory Column Index (IMCI) in vector index memory using and storage empty between using.

• Optimized What is In-Memory Column Index (IMCI) window function execution efficiency, combined and PARTITION BY defining mutual same window function.

Description

• Fixed a potential crash during cost-based query rewrite for SQL statements that contain both an inline view and a scalar subquery.

• Fixed an issue where Index Join intermittently triggered errors when executing partitioned table queries in Hybrid Plan mode.

• Fixed an issue that prevented some statements in prepare statement mode from being correctly routed to column store nodes.

• Fixed What is In-Memory Column Index (IMCI)Delta module order column ization process in rollback row judging breaking legacy missing, causing empty pointer access and triggering cluster crash.

• Fixed an issue where some queries could not be pushed down to In-Memory Column Index for execution when the imci_implicit_orderby_rollup parameter was enabled.

• Fixed an issue that prevented queries with overly long IN-LISTs from using the columnstore index.

Description

• Added the column encryption feature, which works with TDE to encrypt specified columns. A decryption function can be used to retrieve original data when accessing the data.

• Added a feature that deconstructs query expressions with multi-value conditions (IN-LIST) or logical disjunctions (OR) into a UNION ALL operation of multiple independent subqueries to generate index access plans more efficiently.

• Delete statement Added support for Index Hint, i.e. through USE INDEX, IGNORE INDEX syntax specifying specific index using. can through parameter delete_enable_index_hint enabling this feature.

• Lateral horizontal toward derived table (Lateral Derived Table) Added support for parallel execution, in parallel query scenarios support pushed down to Worker execution.

• What is In-Memory Column Index (IMCI) Added support for in subquery in containing has Roll UP in projection column in containing with has aggregation combined function.

• What is In-Memory Column Index (IMCI) Added vector index feature, support base at ANN algorithm near similar retrieval.

Description

• Improved the asynchronous operation and flow control for garbage file cleanup after a read-only IMCI node is deleted.

• Optimized the ALTER TABLE cold storage archiving by removing the requirement to retain one InnoDB partition, thereby supporting archiving of all partitions. Also supports restoring OSS partitions of archived partitioned tables back to InnoDB partitions.

• Optimized What is In-Memory Column Index (IMCI) HASH JOIN structure building Hash table process, HASH table specific implementing, significantly improving What is In-Memory Column Index (IMCI)HASH JOIN performance.

Description

• Fixed a crash that occurred in a parallel query if a constant table was an outer table in a LEFT JOIN with multiple inner tables.

• Fixed an issue where the SCAN command of the Orca feature could fail to return a result, remaining in a searching state.

• Fixed an issue where Sharding Outline did not correctly support table aliases, ORDER BY, and LIMIT BY.

• Fixed a crash caused by the base table and the dependent table not being pushed down to the same shard for execution when a default value references a table.

• Fixed an issue where adding an index to a table with a SET type required a table lock.

• Fixed a cluster crash triggered when you update a NULL value to an empty string in a table that uses the REDUNDANT row format and contains an instant-add column.

• Fixed What is In-Memory Column Index (IMCI) in DISTANCE expression computation DOT mutual similar degree returns value issue.

• Fixed an error that occurred when a cold-archived ORC-format table contained BIT, ENUM, or SET types.

• Fixed INDEX MERGE and index condition under pushing (ICP) plan under adaptive should execution statistics has deviation difference issue.

• Fixed an issue where reading JSON data on a read-only (RO) node could return an INVALID JSON error in some scenarios.

• Fixed a crash caused by an ORDER BY clause being merged into the outer query in a materialized view.

• Fixed an issue where disabling function pushdown caused incorrect results for Elastic Parallel Query.

• Fixed two issues in Data Lifecycle Management (DLM): an archiving exception that left residual temporary tables, for which error logging is now added; and an ASSERT crash caused by an X lock conflict during the second lock upgrade in the DLM partition metadata lock (MDL) archiving process.

• Fixed What is a partitioned table RENAME DDL after tables What is In-Memory Column Index (IMCI) INFORMATION_SCHEMA table in table_name cache not updating issue.

• Fixed an issue in the prepare execute implementation where the collation of string types was incorrectly affected by the character_set_client setting.

• Fixed an assertion error that was triggered by an update task based on old IMCI information after an INSTANT DDL statement was committed.

• Fixed a crash that occurred in high-concurrency scenarios when both Elastic Parallel Query and performance_schema were enabled.

• Fixed an issue where the MySQL Community Edition did not include the LIMIT keyword in query performance cost calculations.

• Fixed an issue where intermittent timing inaccuracies occurred when using imci_analyze_query for query profiling.

• Fixed an issue where unsupported user-defined functions were adaptively routed to IMCI, which caused query errors.

• Fixed a crash in the background thread that collects IMCI statistics for the hybrid optimizer. This bug had a higher probability of being triggered when a virtual column was present.

• Fixed an issue that could occur when accessing a partitioned table after it was restored by using the database and table restoration feature, if the table included an INSTANT ADD COLUMN and had 10 or more partitions.

• Fixed an out-of-memory error that could occur in IMCI when a GROUP BY operation was performed on variable-length strings with a large amount of data.

• Fixed read-only node execution show binlog events error file not existing issue.

• Fixed an issue where queries intermittently could not use Hybrid Plan after enabling global strong consistency (high-performance mode).

• Fixed What is In-Memory Column Index (IMCI) within component thread pool elastic shrinking capacity when producing memory leak issue, avoiding Serverless cluster PCU adjusting when memory leak.

• Fixed What is In-Memory Column Index (IMCI) in loading NCI (Non-Clustered Index) period between execution DROP INDEX operation could cause loading process blocking issue.

• Fixed an issue where a crash occurred before the DDL log for a partition swap was written. This issue caused the cluster's crash recovery process to skip the DDL log, resulting in a cluster anomaly.

• Fixed an issue in parallel query scenarios for partitioned tables where the degree of parallelism for a parallel plan involving a PartitionWise Repartition exceeded max_parallel_degree.

Description

• Fixed a crash in IMCI Hybrid Plans that occurred when using an index join with an anti semi join and a post filter.

• Fixed a system crash during online DDL operations. This crash was caused by data inconsistency when updating or deleting data with a dynamic mask.

• Fixed a crash during data cleanup (TTL) that was caused by a deadlock triggering a transaction rollback.

• Fixed data cleanup (TTL) thread in disabling process in if not releasing transaction (trx) for object, could cause a crash.

• Fixed an Out of Memory (OOM) error that occurred during IMCI sort compaction.

Description

Fixed What is In-Memory Column Index (IMCI) within component thread pool in elastic shrinking capacity process in producing memory leak issue, valid avoiding Serverless cluster in PCU (PolarDB Capacity Unit) dynamic adjusting when occurring memory leak.

Description

• The read-only node (RO) now includes a parse buf (parse log memory block) unused memory reclamation feature to reduce memory usage.

• Window functions now support the frame clause, including the Rows frame and the default Range frame.

• Added support for to Online DDL method modifying What Is X-Engine auto-increment column value.

• Added pruner statistics. You can now query the INFORMATION_SCHEMA.IMCI_SQL_PROFILING view to check the data block pruning status for the TableScan operator.

• You can now use hints to influence the execution plan of statements within a view definition.

• Added support for in INFORMATION_SCHEMA.POLAR_INDEX_STATISTICS view in, query related system table to obtaining each table index named in statistics.

• Added FORCE STORAGE OSS option, Added support for deleting cold data during operations, synchronizing deleting OSS on data.

• Introduced the Range-prefix Skip Scan mechanism to expand the applicability of Skip Scan and improve the performance of complex queries.

• Added In-Memory Column Index (IMCI) for Hybrid Plan feature extending, causing its support not displaying format defining primary key table.

• Added a parameter adaptive_digest_length_multiplier, implementing based on SQL text local length dynamic adjusting Digest pre- allocation memory size feature.

• Added a parameter oss_orc_read_buffer_size, support dynamic configuration ORCReader reading cache size.

• Added a parameter Innodb_ttl_finished_job_expired_days, using at controlling cleaning up ttl_job_history table within expired data.

• In-Memory Column Index (IMCI) in Hybrid Plan Index Join Added support for Semi Join and Anti Semi Join mode support.

Description

• Optimized In-Memory Column Index (IMCI) query performance for IN-List scenarios.

• Optimized high concurrent under using max_execution_time when using the feature, database performance changing difference issue.

• Optimized What is Orca (Redis-Compatible) support HASH data type based on according to field dimension degree setting TTL expired when between.

• Optimized high performance global consistent related changing volume setting from system level alias adjusting as Connection level alias, Added support for database generation in configuration whether enabling.

• Optimized In-Memory Column Index (IMCI) in, Hash Groupby implementing mechanism, implementing based on data minute distributing special feature adaptive should selecting partition strategy and algorithm improving query performance.

Description

• Fixed an issue where the ALTER TABLE command could fail and cause a database restart.

• Fixed an issue with the snapshot feature on column store nodes that could cause monitoring to freeze due to long lock waits.

• Fixed a deadlock caused by concurrent threads while archiving partition data in a partitioned table.

• Fixed an issue where a parallel query with an IS NULL condition could return incorrect results.

• Fixed Adaptive execution plan switching feature in specific type query in scenarios where could cause a cluster crash.

• Fixed In-Memory Column Index (IMCI) multiple instance shared architecture structure under intermittent ImciFollower memory data block accumulation issue.

• Fixed SQL statement holding has table level MDL (metadata lock) but not holding has partition-level MDL lock in cases where, occurring execution DROP PARTITION statement deleting partition, could cause SQL access crash.

• Fixed What Is X-Engine in Index_merge path in processing process in not can correct obtaining Instant_ddl metadata data, causing data query occurring exception.

• Fixed In-Memory Column Index (IMCI) mixed combined materialized Mixoutput read timestamp type data not do time zone converting ization issue.

• Fixed an issue where a column store node occasionally accessed pre-DDL cached data after a DDL operation.

• Fixed an issue where has large amount of field table on performing Instant modify column during operations existing memory access out-of-bounds issue.

• Fixed an issue when enabling asynchronous execution when using the feature, dynamic modifying thread_pool_size parameter could cause database crash.

• Fixed an occasional out-of-bounds read/write issue with ROW LOG when performing an ONLINE DDL operation on a table with a virtual column.

• Fixed string-to-vector type conversion issues by allowing the use of multiple whitespace characters.

• Fixed In-Memory Column Index (IMCI)Serverless cluster in primary node specification than relatively large when unable to shrinking capacity issue.

• Fixed an issue where In-Memory Column Index (IMCI) could not control sorting of query results with ROLLUP through the imci_implicit_orderby_rollup switch.

• Fixed an issue that could cause the database to crash if a SQL query included a GROUP BY WITH ROLLUP clause, the GROUP BY column referenced a column in a derived table, and the optimizer flattened that derived table.

• Fixed an issue that prevented a new read/write node from starting when added to a multi-master cluster with a global read-only node.

• Fixed a potential crash on a secondary node in a multi-master cluster when switching the database or object endpoint.

• Fixed dynamic disabling innodb_polar_log_rdma_transfer parameter after unable to dynamic modifying innodb_log_buffer_size issue.

• Fixed an issue where correlated subqueries containing LIMIT 1 could not be executed on In-Memory Column Index (IMCI).

• Fixed a potential corruption issue with a secondary index created on a virtual column.

Description

• Added an automatic collection feature for column statistics to keep them up to date.

• Added a cost threshold parameter to dynamically control adaptive request distribution between row and column stores. This parameter is distinct from the general cost distribution threshold.

• Added support for RANGE expression partitioning with the cast_for_range() function, which converts double type data to int type, enabling support for double type partition key fields.

• PolarDB for MySQL Added asynchronous execution capability, support asynchronous processing lock and other pending and asynchronous log disk flush and other pending logical.

• Orca provides comprehensive connection monitoring metrics, including total connections, current active connections, and current requests.

• Added explain analyze statement for Outline invoking support, causing its feature and explain statement consistent, improving query Optimized capability.

• Added support for below GIS related function support: ST_HAUSDORFFDISTANCE, ST_FRECHETDISTANCE, ST_LineInterpolatePoints, ST_LineInterpolatePoint, ST_PointAtDistance, ST_Union, ST_Difference, ST_Buffer, ST_Collect, ST_Symdifference, ST_Intersection, enriching GIS feature support.

• Added What is In-Memory Column Index (IMCI) query optimization first level scheduling feature, allowing column storing query allocation optimization first level, Optimized query execution efficiency.

• The TTL (data cleanup) feature added compatibility with TiDB and other products, and added support for the TTL = 'b' + INTERVAL 3 MONTH; syntax.

Description

• Optimized the partition cost model. For equality and IN condition queries on indexes that do not contain a partition key, a partition count factor is included in the cost estimation to improve its accuracy.

• Optimized the algorithm selection logic to allow some correlated subqueries to use an efficient hash join.

• Optimized What is In-Memory Column Index (IMCI) in processing re- miscellaneous IN and EXISTS subquery when performance, meanwhile Improved Optimized for subquery processing logical, thereby significantly improving query efficiency and execution performance.

• Optimized What is Multi-Master Cluster (Limitless) support Partition Redo and asynchronous Redo mechanism, valid improving cursor play in scenarios where system throughput, thereby more good satisfying high concurrent requiring requesting.

• Optimized Orca in processing Redis SCAN command attached with COUNT option when logical, correct maintaining returns metadata element number accurate and consistent.

• Optimized in IMCI in enhancing for SELECT DISTINCT ... statement support, allowing existing non-full-groupby situation, thereby improving and more multiple scenarios compatible with, Optimized query flexible active and suitable using.

• Optimized the logic for multi-level nested IN subqueries, so that they can now be accelerated by the columnstore index.

Description

• Fixed an issue where executing INSERT ON DUPLICATE KEY UPDATE on tables with interval range partitioning could access an invalid memory address.

• Fixed an incompatibility with the database and table restoration feature that occurred because the loose_innodb_initialize_in_space_extend parameter for file space extension could not be set to 0.

• Fixed What is In-Memory Column Index (IMCI) in Hybrid Plan in Index Join mechanism in Added for Semi Join and Anti Semi Join mode support, advancing one step extending query Optimized capability, improving re- miscellaneous query scenarios execution efficiency.

• Fixed a system crash that could occur when using the TempTable engine with a common table expression (CTE) that is referenced multiple times.

• Fixed a query exception that could be triggered when all data in a stored data block was identical.

• Fixed an issue where a read-only node did not sync the gtid_purged variable, which caused queries to return empty values.

• Fixed an issue where a database and table restoration task could be interrupted if the source table had a columnstore index.

• Fixed What is In-Memory Column Index (IMCI) in running row in disabling Hybrid Plan when could cause a crash.

• Fixed a potential socket leak in the Orca network framework when a connection timed out or was actively closed.

• Fixed a rare issue where, with the non-blocking DDL feature enabled, a statement failed to exit normally if its corresponding DDL connection was terminated.

• Fixed an inefficiency in the data lifecycle management (DLM) partition archiving process by using a partition-level metadata lock (MDL) instead of a table-level one, which reduces lock granularity and improves system concurrency and archiving efficiency.

• Fixed inaccurate memory preemption under mixed workloads.

• Fixed an issue where an X-Engine partition required a table rebuild during an optimize operation.

• Fixed a parallel query cluster crash that was caused by a sort operation after a materialized table merge.

• Fixed a potential crash that occurred when the materialized table pruning feature was used with the WITH ROLLUP feature.

• Fixed a cluster crash that occurred when creating a temporary table with a CREATE TABLE AS SELECT statement.

• Fixed a potential cluster crash that could be triggered when accessing statistics tables in a Multi-master Cluster (Limitless) Edition.

• Fixed an index selection error in IMCI queries for constant range filtering scenarios (const BETWEEN field1 AND field2).

• Fixed an issue in Multi-master Cluster (Limitless) Edition scenarios where the online DDL process for the columnstore index was incompatible with the Logical Log Sequence Number (LLSN) logic, causing some columnstore redo log replays to be missed.

• Fixed What Is X-Engine in execution DDL changing more after old tablespace reclaiming leak issue.

• Fixed a crash that was caused by executing an explain for connection statement during a query on an IMCI partitioned table.

• Fixed an issue where IMCI data could not be recovered due to a metadata lock (MDL) acquisition failure during the metadata recovery phase.

• Fixed an issue where queries on IMCI I_S (IMCI-related views in the Information_Schema database) with a name filter condition returned incomplete results because the query did not use a case-sensitive string comparison function.

• Fixed What Is X-Engine read-only node restart when non- pre- period cleaning up data block metadata issue.

• Fixed a potential anomaly in the metadata lock (MDL) system when the number of connections exceeded 65,536.

Description

• Fixed a potential data redundancy issue in the column index that could occur during crash recovery if a column store node cluster restarted during a large transaction.

• Fixed TTL feature syntax issue, would TTL adjusting as non- retaining keyword, to correct maintaining and community region compatible with.

Description

Description

Fixed an issue where reading JSON or GIS large fields through In-Memory Column Index (IMCI) could cause intermittent data read failures.