本文约800字，建议阅读5分钟本文与你分享 Apache IoTDB v0.12.5 的Release Announcement。

Release Announcement

v0.12.5

Apache IoTDB v0.12.5 已经发布，在此版本中，我们修复了一些 Bug，并对内存控制和查询线程模型进行了一些改进。此外，我们在此版本中添加了一个名为"extreme"的聚合函数，并提供了一个用于拆分 TsFile 的工具。

欢迎在我们的网站下载最新版本:

Release Notes

## New Features

* [IOTDB-2078] Split large TsFile tool

* [IOTDB-2192] Support extreme function

## Improvements

* [IOTDB-1297] Refactor the memory control when enabling time partitions

* [IOTDB-2195] Control the concurrent query execution thread

* [IOTDB-2475] Remove sg not ready log in batch process

* [IOTDB-2502] Add query sql in error log if encountering exception

* [IOTDB-2506] Refine the lock granularity of the aggregation query

* [IOTDB-2534] add character support while using double quote

* [IOTDB-2562] Change default value of sync mlog period parameter

* Avoid too many warning logs being printed, when opening too many file handlers

## Bug Fixes

* [IOTDB-1960] Fix count timeseries bug in cluster mode

* [IOTDB-2174] Fix Regexp filter serializing and deserializing error

* [IoTDB-2185] fix get an exception bug when parsing the header of CSV

* [IOTDB-2194] Fix SHOW TIMESERIES will only display 2000 timeseries in cluster mode

* [IOTDB-2197] Fix datatype conversion exception in Spark Connector

* [IOTDB-2209] Fix logback CVE-2021-42550 issue

* [IOTDB-2219] Fix query in-memory data is incorrect in cluster mode

* [IOTDB-2222] Fix OOM and data was written in incorrectly bugs of Spark Connector

* [IOTDB-2251] [IOTDB-2252] Fix Query deadlock

* [IOTDB-2282] fix tag recover bug after tag update

* [IOTDB-2320] MemoryLeak cause by wal Scheduled trim task thread

* [IOTDB-2381] Fix deadlock caused by incorrect buffer pool size counter

* [IOTDB-2400] Fix series reader bug

* [IOTDB-2426] WAL deadlock caused by too many open files

* [IOTDB-2445] Fix overlapped data should be consumed first bug

* [IOTDB-2499] Fix division by zero error when recovering merge

* [IOTDB-2507] Fix NPE when merge recover

* [IOTDB-2528] Fix MLog corruption and recovery bug after killing system

* [IOTDB-2532] Fix query with align by device can't get value after clear cache

* [IOTDB-2533] Fix change max_deduplicated_path_num does not take effect

* [IOTDB-2544] Fix tag info sync error during metadata sync

* [IOTDB-2550] Avoid show timeseries error after alter tag on sync sender

* [IOTDB-2567] Fix thread and ByteBuffer leak after service stopped

* [IOTDB-2568] "show query processlist" is blocked

* [IOTDB-2580] Fix DirectByteBuffer and thread leak when deleting storage group

* [IOTDB-2584] Fix cross space compaction selector

* [IOTDB-2603] Fix compaction recover

* [IOTDB-2604] Fix batch size is invalid in import-csv tool

* [IOTDB-2620] Unrecognizable operator type (SHOW) for AuthorityChecker

* [IOTDB-2624] Fix "overlapped data should be consumed first" occurs when executing query

* [IOTDB-2640] Fix cross compaction recover bug

* [IOTDB-2641] Fix time range of TsFile resource overlaps after unseq compaction

* [IOTDB-2642] Fix the new file has a higher compact priority than the old file in unseq compaction

* Fix a logical bug in processPlanLocally in cluster mode

* Throw Exception while using last query with align by device

* Add a judgement to determine raft log size can fit into buffer before log appending in cluster mode

* Fix grafana can't be used bug