10.1.x Resolved issues¶
10.1.1-26¶
Release Date: 12/20/2022
| Issue Key | Component | Description |
|---|---|---|
| VER-84175 | Control Networking | Previously, addition of standby nodes changed the assignment of control nodes and number of fault groups. This issue has been resolved: standby nodes are no longer set as control nodes. 以前,添加备用节点会改变控制节点的分配和故障组数量。此问题已解决:备用节点不再被设置为控制节点。 |
| VER-84242 | Optimizer | Queries with string expressions in SELECT statements that were part of other expressions and also contained constant expressions sometimes returned an error. The issue has been resolved. SELECT 语句中包含字符串表达式(这些表达式是其他表达式的一部分并同时包含常量表达式)的查询有时返回错误。此问题已解决。 |
10.1.1-25¶
Release Date: 11/7/2022
| Issue Key | Component | Description |
|---|---|---|
| VER-83752 | Backup/DR | VBR now retries S3 uploads on receiving failure messages instead of returning errors. VBR 现在在收到失败消息时会重试 S3 上传,而不是直接返回错误。 |
| VER-83923 | Optimizer | Access policy rules on UPDATE of a table column are now applied only once instead of twice. UPDATE 表列时的访问策略规则现在只应用一次,而不是两次。 |
10.1.1-23¶
Release Date: 09/12/2022
| Issue Key | Component | Description |
|---|---|---|
| VER-83257 | Execution Engine | COUNT with complex type arguments in versions <11.0.x previously caused crashes; now returns an error. 在低于 11.0.x 的版本中,使用复杂类型参数的 COUNT 此前会导致崩溃;现在会返回错误。 |
| VER-83291 | Data load / COPY | Reduced CPU usage for operations over directories with large numbers of external files on object stores. 减少了对象存储上包含大量外部文件的目录操作中的 CPU 使用率。 |
10.1.1-22¶
Release Date: 08/24/2022
| Issue Key | Component | Description |
|---|---|---|
| VER-82761 | Execution Engine, Optimizer | Merge join inputs now maintain necessary sort keys when sort keys were not fully maintained with multiple equivalent sort keys, fixing incorrect results. 当多个等效排序键未完全维护时,合并连接输入现在会维护必要的排序键,修复了错误结果。 |
| VER-82807 | Backup/DR | Restoring a backup to a database with a different communal storage location no longer fails if the database's oid was assigned to another object. 如果数据库的 oid 已分配给另一个对象,将备份恢复到具有不同公共存储位置的数据库不再失败。 |
10.1.1-21¶
Release Date: 08/01/2022
| Issue Key | Component | Description |
|---|---|---|
| VER-82379 | Admin Tools | Adding two nodes to a one-node enterprise database and rebalancing no longer fails in admintool CLI or GUI. 向单节点企业数据库添加两个节点并重新平衡不再在 admintool CLI 或 GUI 中失败。 |
| VER-82631 | Client Drivers - VSQL | vsql now returns an error code when encountering errors during long queries (e.g., exceeding RUNTIMECAP). vsql 现在在长查询(例如超过 RUNTIMECAP)期间遇到错误时会返回错误码。 |
| VER-82776 | Execution Engine | The makeutf8 function no longer causes undefined behavior with maximum-length inputs. makeutf8 函数在最大长度输入下不再导致未定义行为。 |
10.1.1-20¶
Release Date: 05/31/2022
| Issue Key | Component | Description |
|---|---|---|
| VER-81740 | UI - Management Console | Cookie features are now set properly in HTTP headers. Cookie 功能现在在 HTTP 标头中正确设置。 |
| VER-81777 | Optimizer | DBD functions DESIGNER_ADD_DESIGN_QUERIES and DESIGNER_RUN_POPULATE_DESIGN_AND_DEPLOY now support communal storage locations in addition to local file system paths. DBD 函数 DESIGNER_ADD_DESIGN_QUERIES 和 DESIGNER_RUN_POPULATE_DESIGN_AND_DEPLOY 现在除了本地文件系统路径外,还支持公共存储位置。 |
| VER-81861 | ResourceManager | Disk space tracking no longer initializes incorrectly during node startup, eliminating spurious 'Spurious DiskSpaceWritten' messages. 磁盘空间跟踪在节点启动期间不再错误初始化,消除了虚假的 'Spurious DiskSpaceWritten' 消息。 |
10.1.1-19¶
Release Date: 05/4/2022
| Issue Key | Component | Description |
|---|---|---|
| VER-81270 | Data Collector | Setting an empty channel in SET_DATA_COLLECTOR_NOTIFY_POLICY is now disallowed. 现在禁止在 SET_DATA_COLLECTOR_NOTIFY_POLICY 中设置空通道。 |
| VER-81387 | Backup/DR | VBR no longer errors when connecting to Vertica through a non-primary subscriber node during Eon replication tasks. VBR 在 Eon 复制任务期间通过非主要订阅者节点连接到 Vertica 时不再报错。 |
| VER-81447 | DDL - Projection | Projections in Eon mode no longer have a _b# suffix.Eon 模式下的投影不再带有 _b# 后缀。 |
10.1.1-18¶
Release Date: 04/07/2022
| Issue Key | Component | Description |
|---|---|---|
| VER-81145 | Admin Tools | admintools -t re_ip options -T and -U now correctly update admintools.conf with the proper control messaging protocol. admintools -t re_ip 的 -T 和 -U 选项现在使用正确的控制消息协议正确更新 admintools.conf。 |
| VER-81049 | Execution Engine | Event series joins now use the correct tuple schema when initializing cached rows with nulls, preventing crashes. 事件序列连接在初始化带有空值的缓存行时现在使用正确的元组架构,防止崩溃。 |
| VER-81045 | UI - Management Console | MC now loads all data when a custom file system is detected during AWS cluster scaling, and the AWS Key Pair populates correctly. MC 现在在 AWS 集群扩展期间检测到自定义文件系统时会加载所有数据,并且 AWS Key Pair 会正确填充。 |
| VER-80907 | UI - Management Console | MC now saves a DBD-generated design regardless of the K-safety setting (previously required K-safety=0). MC 现在无论 K-safety 设置如何都会保存 DBD 生成的设计(此前需要 K-safety=0)。 |
| VER-80902 | UI - Agent | HSTS is now enabled in the agent APIs for HTTPS connections. 代理 API 中现在为 HTTPS 连接启用了 HSTS。 |
| VER-80744 | UI - Management Console | Insecure HTTP responses are now prevented when a non-integer dbid is sent to /webui/databases/{dbid}.当向 /webui/databases/{dbid} 发送非整数 dbid 时,现在会阻止不安全的 HTTP 响应。 |
10.1.1-17¶
Release Date: 03/17/2022
| Issue Key | Component | Description |
|---|---|---|
| VER-80225 | UI - Management Console | The analytics chart selected from the toolbar dropdown on the Activities tab now displays correctly. 从 Activities 选项卡的工具下拉菜单中选择的分析图表现在正确显示。 |
| VER-80841 | Execution Engine | IN expression evaluation for cases where the right-hand side is neither a subquery nor a constant list no longer crashes due to flawed detection. IN 表达式求值在右侧既不是子查询也不是常量列表的情况下,不再因检测缺陷而崩溃。 |
| VER-80848 | AP-Geospatial | STV_Create_Index no longer creates incorrect indexes on large polygon sets that could cause query failures or node crashes. STV_Create_Index 不再在大型多边形集上创建可能导致查询失败或节点崩溃的错误索引。 |
| VER-80896 | Backup/DR, Migration Tool | Loading data into tables with unsegmented projections after migrating from Enterprise to Eon mode no longer fails with "Cannot plan query because no super projections are safe." 从 Enterprise 模式迁移到 Eon 模式后,向具有非分段投影的表加载数据不再失败并提示 "Cannot plan query because no super projections are safe."。 |
| VER-80899 | Admin Tools | db_add_node now uses any UP node as the source for syncing vertica.conf and spread.conf instead of requiring the first node. db_add_node 现在使用任意 UP 节点作为同步 vertica.conf 和 spread.conf 的来源,而不是要求第一个节点。 |
10.1.1-16¶
Release Date: 02/28/2022
| Issue Key | Component | Description |
|---|---|---|
| VER-79876 | UI - Management Console | The KPI dropdown on the Manage page now opens correctly, and the Activity Page no longer displays as unauthorized. Manage 页面上的 KPI 下拉菜单现在正确打开,Activity 页面不再显示为未授权。 |
| VER-80452 | Execution Engine, UDX | RIGHT is no longer an inline SQL function — now defined internally — resolving errors with volatile UDA parameters. RIGHT 不再是一个内联 SQL 函数——现在在内部定义——解决了易变 UDA 参数的错误。 |
| VER-80541 | Backup/DR | Error message for CREATE TABLE AS during replication now clearly indicates the source of the problem. 复制期间 CREATE TABLE AS 的错误消息现在清晰指示问题来源。 |
| VER-80601 | Optimizer, Performance tests | Running analyze_statistics('') on large catalogs no longer runs out of memory.在大型目录上运行 analyze_statistics('') 不再耗尽内存。 |
| VER-80648 | DDL | IMPORT_STATISTICS can now import statistics from EXPORT_STATISTICS_PARTITION output. IMPORT_STATISTICS 现在可以从 EXPORT_STATISTICS_PARTITION 输出导入统计信息。 |
| VER-80739 | Execution Engine | regexp_like() no longer returns incorrect results when regex contains zero-match patterns like "?" and "". regexp_like() 在正则表达式包含零匹配模式(如 "?" 和 "")时不再返回错误结果。 |
10.1.1-15¶
Release Date: 02/04/2022
| Issue Key | Component | Description |
|---|---|---|
| VER-79882 | Tuple Mover | MaxDVROSPerContainer knob now accepts any value >1 with a formula of max(2, (MaxDVROSPerContainer+1)/2) to avoid excessive DVMergeouts.MaxDVROSPerContainer 参数现在接受任何大于 1 的值,使用公式 max(2, (MaxDVROSPerContainer+1)/2) 来避免过多的 DVMergeout。 |
| VER-79954 | Depot | copy_partitions_to_table on two tables with identical pinning policies no longer crashes when the target has no projection data.在两个具有相同固定策略的表上执行 copy_partitions_to_table,当目标没有投影数据时不再崩溃。 |
| VER-79961 | UI - Management Console | The MC feedback form no longer causes slow loading of other MC pages. MC 反馈表单不再导致其他 MC 页面加载缓慢。 |
| VER-80235 | Execution Engine | Queries comparing numeric values to non-numeric strings like 'abc' no longer fail due to runtime predicate reordering. 将数值与非数字字符串(如 'abc')进行比较的查询不再因运行时谓词重排序而失败。 |
| VER-80239 | Recovery | After rebalance, Tuple Mover now groups ROS containers in batches smaller than MaxMrgOutROSSizeMB; larger containers merge individually. 重新平衡后,Tuple Mover 现在将小于 MaxMrgOutROSSizeMB 的 ROS 容器分批分组;较大的容器单独合并。 |
| VER-80372 | DDL, Execution Engine | Server no longer crashes when system queries use large strings as predicate constants (e.g., table_name = ...).当系统查询使用大字符串作为谓词常量(例如 table_name = ...)时,服务器不再崩溃。 |
| VER-80525 | ComplexTypes, Execution Engine | "subquery." as an expression argument now expands into a complex ROW expression with one field per subquery column, preventing undefined behavior. "subquery." 作为表达式参数现在扩展为一个复杂的 ROW 表达式,每个子查询列对应一个字段,防止未定义行为。 |
10.1.1-14¶
Release Date: 01/12/2022
| Issue Key | Component | Description |
|---|---|---|
| VER-80171 | Kafka Integration | log4j library updated to resolve CVE-2021-45105 and CVE-2021-44832. log4j 库已更新以解决 CVE-2021-45105 和 CVE-2021-44832。 |
| VER-80243 | UI - Management Console | log4j library updated to resolve CVE-2021-44832. log4j 库已更新以解决 CVE-2021-44832。 |
10.1.1-13¶
Release Date: 12/17/2021
| Issue Key | Component | Description |
|---|---|---|
| VER-79222 | UI - Management Console | MC now correctly imports databases containing a table named "users" in the public schema. MC 现在可以正确导入包含 public 模式中名为 "users" 的表的数据库。 |
| VER-79622 | Optimizer | The optimizer now correctly uses available LAP/Top-K projections when directed to do so. 优化器现在在被指示时正确使用可用的 LAP/Top-K 投影。 |
| VER-79668 | FlexTable | MapToString no longer returns null when the underlying VMap was not null (display issue only, no data loss). MapToString 在底层 VMap 不为空时不再返回 null(仅显示问题,无数据丢失)。 |
| VER-80083 | Kafka Integration | log4j library updated to resolve CVE-2021-44228 and CVE-2021-45046. log4j 库已更新以解决 CVE-2021-44228 和 CVE-2021-45046。 |
| VER-80085 | UI - Management Console | log4j library updated to resolve CVE-2021-44228 and CVE-2021-45046. log4j 库已更新以解决 CVE-2021-44228 和 CVE-2021-45046。 |
10.1.1-12¶
Release Date: 11/26/2021
| Issue Key | Component | Description |
|---|---|---|
| VER-79841 | Data Removal - Delete, Purge, Partitioning, EON | Eon mode now supports optimized delete and merge for unsegmented projections. Eon 模式现在支持非分段投影的优化删除和合并。 |
10.1.1-11¶
Release Date: 11/24/2021
| Issue Key | Component | Description |
|---|---|---|
| VER-79517 | Execution Engine | Database no longer crashes when query plans with filter operators are canceled. 当带有过滤运算符的查询计划被取消时,数据库不再崩溃。 |
10.1.1-10¶
Release Date: 10/27/2021
| Issue Key | Component | Description |
|---|---|---|
| VER-78240 | ComplexTypes, Kafka Integration | Data loads with array data types no longer cause failures in fenced mode filter or parser. 具有数组数据类型的数据加载不再在隔离模式过滤器或解析器中导致失败。 |
| VER-79226 | Data Removal - Delete, Purge, Partitioning, DDL | SWAP_PARTITIONS_BETWEEN_TABLES now ignores metadata differences in epoch columns after upgrading from 9.3.1. 从 9.3.1 升级后,SWAP_PARTITIONS_BETWEEN_TABLES 现在忽略 epoch 列中的元数据差异。 |
| VER-79245 | DDL - Projection, Optimizer | Eligible predicates are now pushed down into subqueries with a LIMIT OVER clause. 符合条件的谓词现在可以被下推到带有 LIMIT OVER 子句的子查询中。 |
| VER-79250 | Hadoop | Parquet export row limit raised from 2^31 to 2^64 rows to prevent node failures. Parquet 导出行限制从 2^31 提高到 2^64 行以防止节点故障。 |
| VER-79252 | Execution Engine | Memory accumulated by sessions using User Defined Load code on non-initiator nodes is now properly released. 在非发起者节点上使用用户定义加载代码的会话所累积的内存现在被正确释放。 |
| VER-79415 | Tuple Mover | DVMergeout now calculates column offset and length instead of accessing storage files, reducing S3 API calls. DVMergeout 现在计算列偏移量和长度,而不是访问存储文件,减少了 S3 API 调用。 |
10.1.1-9¶
Release Date: 10/12/2021
| Issue Key | Component | Description |
|---|---|---|
| VER-78905 | License | Background license size verifier no longer runs excessively long in Eon mode. 后台许可证大小验证器在 Eon 模式下不再运行过长的时间。 |
| VER-78946 | UI - Management Console | MC no longer errors when configuring email gateway aliases containing hyphens. MC 在配置包含连字符的电子邮件网关别名时不再报错。 |
| VER-78987 | Client Drivers - ADO | Canceled queries with immediate DataReader.close() no longer propagate exceptions incorrectly. 取消查询后立即调用 DataReader.close() 不再错误传播异常。 |
| VER-78989 | Optimizer | Adding a column with a default containing a TIMESERIES clause no longer returns a ROLLBACK. 添加包含 TIMESERIES 子句的默认值的列不再返回 ROLLBACK。 |
| VER-79043 | Optimizer | Partition statistics now support partition expressions using date_trunc().分区统计信息现在支持使用 date_trunc() 的分区表达式。 |
| VER-79046 | Optimizer | Queries with multiple distinct aggregates no longer produce wrong results when inputs appear segmented on same columns. 当输入在相同列上分段时,具有多个不同聚合函数的查询不再产生错误结果。 |
| VER-79066 | Kafka Integration | Filters like KafkaInsertDelimiters no longer return VIAssert errors with small messages. 像 KafkaInsertDelimiters 这样的过滤器在处理小消息时不再返回 VIAssert 错误。 |
| VER-79230 | UI - Management Console | The feedback feature now sends information by email instead of failing to upload. 反馈功能现在通过电子邮件发送信息,而不是上传失败。 |
10.1.1-8¶
Release Date: 9/23/2021
| Issue Key | Component | Description |
|---|---|---|
| VER-78352 | UI-Management Console | Internet Explorer 11 now displays Queries chart information on the Activity tab. Internet Explorer 11 现在在 Activity 选项卡上显示查询图表信息。 |
| VER-78587 | Database Designer Core | Database Designer no longer generates files with wrong permissions. Database Designer 不再生成具有错误权限的文件。 |
| VER-78664 | Execution Engine | Queries on EXTERNAL_TABLE_DETAILS with complex predicates no longer return wrong results or crash the cluster. 对 EXTERNAL_TABLE_DETAILS 使用复杂谓词的查询不再返回错误结果或导致集群崩溃。 |
| VER-78669 | Backup/DR | Full restore no longer overwrites AWS configuration parameters (AWSEndpoint, AWSAuth, AWSEnableHttps) with backup settings. 完全恢复不再使用备份设置覆盖 AWS 配置参数(AWSEndpoint、AWSAuth、AWSEnableHttps)。 |
| VER-78713 | Optimizer | Memory allocated to query tree objects is now freed when no longer in use, addressing excessive memory consumption. 分配给查询树对象的内存在不再使用时被释放,解决了内存消耗过多的问题。 |
| VER-78727 | Optimizer | The optimizer now properly handles combined NotNull predicates during predicate pushdown. 优化器现在在谓词下推期间正确处理组合的 NotNull 谓词。 |
| VER-78748 | AP-SparkIntegration, Client Drivers - JDBC | V1 Spark Connector no longer exposes passwords and credentials in Spark logs. V1 Spark Connector 不再在 Spark 日志中暴露密码和凭据。 |
10.1.1-7¶
Release Date: 8/19/2021
| Issue Key | Component | Description |
|---|---|---|
| VER-78347 | Data Networking | Socket accepting internal connections no longer erroneously closes and floods vertica.log with errors. 接受内部连接的套接字不再错误关闭并用错误消息淹没 vertica.log。 |
| VER-78350 | S3, Storage and Access Layer | S3 storage Location_path values are now stored in the same case format they were created in. S3 存储的 Location_path 值现在以创建时相同的大小写格式存储。 |
| VER-78351 | UI-Management Console | MC Overview page now displays properly after automatic AWS upgrade. MC Overview 页面在自动 AWS 升级后现在正常显示。 |
| VER-78354 | Catalog Engine | Standby nodes now receive changes to the GENERAL resource pool when replacing a down node. 备用节点在替换故障节点时现在接收 GENERAL 资源池的更改。 |
| VER-78357 | Optimizer | Updating a query result set using volatile function LISTAGG no longer causes unequal row counts among projections. 使用易变函数 LISTAGG 更新查询结果集不再导致投影之间的行数不相等。 |
| VER-78362 | UI-Management Console; Management Console | COPY type queries now appear in Completed Queries on the Query Monitoring Activity page. COPY 类型查询现在出现在 Query Monitoring Activity 页面的 Completed Queries 中。 |
| VER-78364 | ILM, Tuple Mover | Force-split operations with partition management now complete both stages as a single transaction, preventing premature Tuple Mover merge attempts. 带分区管理的强制拆分操作现在将两个阶段作为单个事务完成,防止 Tuple Mover 过早尝试合并。 |
| VER-78441 | Admin Tools | create_db prompt now times out after 120 seconds if the user doesn't respond, then exits. create_db 提示现在在用户未响应时 120 秒后超时退出。 |
| VER-78500 | Installation Program | Vertica now properly handles nofile and nproc values in environments with >1TB RAM. Vertica 现在在内存大于 1TB 的环境中正确处理 nofile 和 nproc 值。 |
| VER-78523 | Depot | Vertica now properly detects and handles corrupted depot index files. Vertica 现在正确检测和处理损坏的 depot 索引文件。 |
10.1.1-6¶
Release Date: 7/21/2021
| Issue Key | Component | Description |
|---|---|---|
| VER-77828 | Backup/DR | STDOUT during object restore in coexist mode now contains the correct schema name prefix. 共存模式中对象恢复期间的 STDOUT 现在包含正确的模式名称前缀。 |
| VER-77871 | Kafka Integration | The Kafka scheduler no longer hangs in rare cases. Kafka 调度器在极少数情况下不再挂起。 |
| VER-77875 | Backup/DR | Access permission check for S3 bucket root during backup/restore removed; users with access to specific directories can now proceed. 移除了备份/恢复期间对 S3 存储桶根目录的访问权限检查;拥有特定目录访问权限的用户现在可以继续操作。 |
| VER-78124 | Catalog Engine | Vertica now restarts properly for nodes with very large checkpoint files. Vertica 现在对于具有非常大检查点文件的节点可以正常重启。 |
| VER-78129 | Optimizer | Views can now be queried with the TABLESAMPLE option. 现在可以使用 TABLESAMPLE 选项查询视图。 |
| VER-78130 | Execution Engine | Query predicates can no longer be reordered unsafely when leading sort order columns are restricted to constant values and not run-length encoded. 当引导排序顺序列被限制为常量值且未进行游程编码时,查询谓词不能再被不安全地重排序。 |
| VER-78132 | Installation Program | The installer now caches get_disks information instead of calling it repeatedly.安装程序现在缓存 get_disks 信息,而不是重复调用它。 |
| VER-78147 | Data Removal - Delete, Purge, Partitioning | Dropping local storage locations for USER data on clusters with standby nodes no longer errors. 在具有备用节点的集群上删除 USER 数据的本地存储位置不再报错。 |
| VER-78201 | Spread | Spread now responds properly to network changes. Spread 现在正确响应网络变化。 |
| VER-78211 | Catalog Engine | Calling close_session() during analyze_statistics() on a local temporary table no longer crashes Vertica.在本地临时表上执行 analyze_statistics() 期间调用 close_session() 不再导致 Vertica 崩溃。 |
10.1.1-5¶
Release Date: 7/13/2021
| Issue Key | Component | Description |
|---|---|---|
| VER-77844 | Kafka Integration | The vkconfig scheduler now properly supports an offset of -3. vkconfig 调度器现在正确支持偏移量 -3。 |
| VER-78065 | Execution Engine | The throw_error function is now marked immutable, so SQL macros can call it multiple times.throw_error 函数现在被标记为不可变,因此 SQL 宏可以多次调用它。 |
| VER-78092 | SDK-Python | Python users can now define scalar functions as immutable. Python 用户现在可以将标量函数定义为不可变。 |
10.1.1-4¶
Release Date: 7/7/2021
| Issue Key | Component | Description |
|---|---|---|
| VER-76256 | Hadoop | Vertica now properly handles queries of external tables of Parquet files with ZSTD compression. Vertica 现在正确处理使用 ZSTD 压缩的 Parquet 文件外部表的查询。 |
| VER-77571 | Data Collector | Custom data collector settings are now properly handled when upgrading to version 10.x. 升级到版本 10.x 时,自定义数据收集器设置现在被正确处理。 |
| VER-77634 | Data Export | A more helpful warning message now displays when using the deprecated s3export function on Minio storage. 在 Minio 存储上使用已弃用的 s3export 函数时,现在显示更有帮助的警告消息。 |
| VER-77928 | Hadoop | Exporting data to Parquet now properly respects resource pool limits. 导出数据到 Parquet 现在正确遵循资源池限制。 |
| VER-77946 | Execution Engine | Fixed incorrect null-block max value check that mistakenly pruned rows from predicate evaluation. 修复了错误的空块最大值检查,该检查错误地修剪了谓词求值中的行。 |
10.1.1-3¶
Release Date: 6/15/2021
| Issue Key | Component | Description |
|---|---|---|
| VER-77181 | Spread | Vertica now properly detects duplicate tokens with the same arq ID. Vertica 现在正确检测具有相同 arq ID 的重复令牌。 |
| VER-77389 | Installation Program | The pre-upgrade script now displays more helpful error messages for version 10+ databases. 预升级脚本现在为版本 10+ 的数据库显示更有帮助的错误消息。 |
| VER-77391 | Data load / COPY, Security | COPY REJECTED DATA to TABLE now properly distributes data between identical table names in different schemas. COPY REJECTED DATA to TABLE 现在在不同模式中相同表名称之间正确分发数据。 |
| VER-77458 | Optimizer | CASE expressions with constant conditions and shorter string results no longer produce internal errors in joins with aggregation. 具有常量条件和较短字符串结果的 CASE 表达式在带聚合的连接中不再产生内部错误。 |
| VER-77490 | Execution Engine | 'count(s.)' with subquery s now properly handles multi-valued expression expectations. 带有子查询 s 的 'count(s.)' 现在正确处理多值表达式预期。 |
| VER-77523 | Security | Vertica now automatically creates default key projections for a user with DML access performing INSERT into a table with a primary key and no projections. Vertica 现在为具有 DML 访问权限的用户向具有主键但无投影的表执行 INSERT 时自动创建默认键投影。 |
| VER-77573 | Monitoring | Startup.log now identifies nodes that were slow to join a cluster. Startup.log 现在标识了加入集群较慢的节点。 |
10.1.1-2¶
Release Date: 6/9/2021
| Issue Key | Component | Description |
|---|---|---|
| VER-77369 | Hadoop | Vertica now connects to HDFS nameservice regardless of upper/lowercase spelling. Vertica 现在可以连接到 HDFS nameservice,无论大小写拼写如何。 |
| VER-77295 | Backup/DR | VBR can now delete restore points that are already partially deleted. VBR 现在可以删除已经部分删除的恢复点。 |
| VER-77246 | Data Removal - Delete, Purge, Partitioning | DML performance improved when loading data into partitioned tables. 将数据加载到分区表时,DML 性能得到改善。 |
| VER-77180 | Execution Engine | Queries on v_internal tables with equality predicates on relname or nspname (including embedded underscore/percent) are now optimized as expected. 对 v_internal 表使用 relname 或 nspname 上的相等谓词(包括嵌入的下划线/百分号)的查询现在按预期进行了优化。 |
| VER-77154 | Tuple Mover | Tuple Mover now considers resource pool capacity when creating a mergeout plan, adjusting ROS container count accordingly. Tuple Mover 现在在创建 mergeout 计划时考虑资源池容量,相应调整 ROS 容器数量。 |
10.1.1-1¶
Release Date: 5/13/2021
| Issue Key | Component | Description |
|---|---|---|
| VER-76899 | Optimizer | NO HISTOGRAM event is no longer incorrectly set on the hidden epoch column of dc_optimizer_events. NO HISTOGRAM 事件不再被错误设置在 dc_optimizer_events 的隐藏 epoch 列上。 |
| VER-76971 | Backup/DR | Object restore now functions properly for objects >1 TB on Eon Mode clusters on S3. 对象恢复现在对 S3 上 Eon 模式集群中大于 1 TB 的对象正常运行。 |
| VER-76994 | Optimizer | Using unary operators as GROUP BY or ORDER BY elements in WITH clause statements no longer crashes Vertica. 在 WITH 子句语句中使用一元运算符作为 GROUP BY 或 ORDER BY 元素不再导致 Vertica 崩溃。 |
| VER-77020 | Client Drivers - ADO | ADO.NET driver now properly cancels active queries instead of running them to completion. ADO.NET 驱动现在正确取消活动查询,而不是将其运行完成。 |
| VER-77021 | Optimizer | Views invoking UDx functions with '' argument no longer crash when the underlying table changes. 使用 '' 参数调用 UDx 函数的视图在底层表更改时不再崩溃。 |
| VER-77079 | Machine Learning | Combined use of the ROC function and predict_svm_classifier is now properly handled. ROC 函数和 predict_svm_classifier 的组合使用现在被正确处理。 |
10.1.1-0¶
Release Date: 4/30/2021
| Issue Key | Component | Description |
|---|---|---|
| VER-76360 | Catalog Sync and Revive, Depot, EON | Secondary subcluster nodes no longer lose depot files on restart. 辅助子集群节点在重启时不再丢失 depot 文件。 |
| VER-76361 | Execution Engine | REGEXP_REPLACE now properly accepts and returns LONG VARCHAR values. REGEXP_REPLACE 现在正确接受和返回 LONG VARCHAR 值。 |
| VER-76400 | Kafka Integration | Kafka UDx now officially supports SASL SCRAM-SHA-256 and SCRAM-SHA-512. Kafka UDx 现在正式支持 SASL SCRAM-SHA-256 和 SCRAM-SHA-512。 |
| VER-76457 | Data Export | Parquet export now correctly writes pre-2000 timestamp time portions as positive numbers, fixing Impala compatibility. Parquet 导出现在正确将 2000 年之前的时间戳时间部分写为正数,修复了 Impala 兼容性。 |
| VER-76475 | Execution Engine | Complex memory-intensive queries on high-core-count machines no longer crash when EXECUTIONPARALLELISM is set high. 在高核心数机器上,复杂内存密集型查询在 EXECUTIONPARALLELISM 设置较高时不再崩溃。 |
| VER-76486 | Depot | System tables no longer report depot files as existing after eviction, preventing "File not found" errors. 系统表在逐出后不再报告 depot 文件存在,防止 "File not found" 错误。 |
| VER-76522 | Installation Program, UI-Management Console | MC installer storage footprint has been optimized. MC 安装程序存储占用已优化。 |
| VER-76581 | Recovery | DFS file handling improved: missing files no longer push recovery epoch behind AHM, and ASR now recovers missing DFS files from replicas or drops unrecoverable ones. DFS 文件处理改进:丢失文件不再将恢复 epoch 推后到 AHM 之后,ASR 现在从副本恢复丢失的 DFS 文件或删除不可恢复的文件。 |
| VER-76631 | Data Export, Optimizer | Write permission checks during Parquet export now only occur on participating nodes. Parquet 导出期间的写权限检查现在只在参与节点上执行。 |
| VER-76655 | UI-Management Console | The k-safety warning dialog's OK button in MC Database Designer wizard mode is now responsive. MC Database Designer 向导模式中 k-safety 警告对话框的 OK 按钮现在可以响应。 |
| VER-76700 | Admin Tools | Adding a node to Eon-mode database no longer prompts to rebalance (unsupported in Eon). 向 Eon 模式数据库添加节点不再提示重新平衡(Eon 中不支持)。 |
| VER-76756 | Optimizer, Security | Users no longer need USAGE privilege on schema B to SELECT view A.v1 when it references base table B.t1. 当视图 A.v1 引用基表 B.t1 时,用户不再需要对模式 B 的 USAGE 权限来 SELECT 该视图。 |
| VER-76769 | Execution Engine | Variable-length-optimized joins no longer risk database crash due to NULL join key checking. 变长优化连接不再因 NULL 连接键检查而有数据库崩溃风险。 |
| VER-76802 | Client Drivers - JDBC, ComplexTypes | JDBC driver now returns the correct type ID for complex types. JDBC 驱动现在为复杂类型返回正确的类型 ID。 |
| VER-76803 | Client Drivers - ODBC, VSQL, OLEDB, Security | FIPS support is restored. FIPS 支持已恢复。 |
10.1.0-3¶
Release Date: 3/17/2021
| Issue Key | Component | Description |
|---|---|---|
| VER-76041 | EON, Tuple Mover | Commit resync is now disabled for background Tuple Mover jobs. 提交重新同步现在对后台 Tuple Mover 作业禁用。 |
| VER-76046 | Execution Engine | Memory used by repeated access privilege checks during export_objects is now freed promptly, preventing OOM crashes.export_objects 期间重复访问权限检查使用的内存现在被及时释放,防止 OOM 崩溃。 |
| VER-76059 | Tuple Mover | The Tuple Mover now always checks for outstanding MERGEOUT requests after completing DVMERGEOUT jobs. Tuple Mover 现在在完成 DVMERGEOUT 作业后始终检查未完成的 MERGEOUT 请求。 |
| VER-76061 | Monitoring | Output of get_expected_recovery_epoch is no longer truncated in clusters with >20 nodes.get_expected_recovery_epoch 的输出在超过 20 个节点的集群中不再被截断。 |
| VER-76136 | Cloud - Azure | Loads on large Azure clusters no longer fail due to storage service throttling. 大型 Azure 集群上的加载不再因存储服务限流而失败。 |
| VER-76137 | Backup/DR | VBR now supports restores from MinIO S3 locations containing special characters. VBR 现在支持从包含特殊字符的 MinIO S3 位置进行恢复。 |
| VER-76190 | Depot | System tables no longer report depot files as existing after eviction (same fix as VER-76486 in later release). 系统表在逐出后不再报告 depot 文件存在(与后续版本中的 VER-76486 相同修复)。 |
| VER-76252 | Data Export | Parquet files exported by Vertica now properly display date values when imported to Impala. Vertica 导出的 Parquet 文件在导入到 Impala 时现在正确显示日期值。 |
| VER-76393 | Data Collector, EON | DC table files are now appropriately cleared when dc sync is disabled. 当 dc 同步禁用时,DC 表文件现在被适当清除。 |
10.1.0-2¶
Release Date: 2/23/2021
| Issue Key | Component | Description |
|---|---|---|
| VER-75827 | Hadoop | Vertica server no longer crashes while logging HDFS events in certain configurations. Vertica 服务器在特定配置下记录 HDFS 事件时不再崩溃。 |
| VER-75901 | Tuple Mover | dc_mergeout_requests now contains valid transaction IDs instead of zero when Mergeout Cache is enabled.当 Mergeout Cache 启用时, dc_mergeout_requests 现在包含有效的事务 ID 而不是零。 |
| VER-75913 | Tuple Mover | Tuple Mover no longer logs excessive PURGE requests on a projection while another MERGEOUT runs on the same projection. 当另一个 MERGEOUT 在同一投影上运行时,Tuple Mover 不再记录过多的 PURGE 请求。 |
| VER-75973 | Hadoop | Fixed Parquet Predicate Pushdown bug where correct rowgroups were not pruned based on predicates when the file was on HDFS. 修复了 Parquet 谓词下推错误,该错误导致文件在 HDFS 上时无法根据谓词修剪正确的行组。 |
10.1.0-1¶
Release Date: 2/9/2021
| Issue Key | Component | Description |
|---|---|---|
| VER-75904 | Security | Vertica no longer takes a catalog lock during authentication after password algorithm changes from MD5 to SHA512. 在密码算法从 MD5 更改为 SHA512 后,Vertica 在身份验证期间不再获取目录锁。 |
10.1.0-0¶
Release Date: 2/1/2021
| Issue Key | Component | Description |
|---|---|---|
| VER-36888 | Admin Tools | stop_db no longer returns early and now returns the proper return code.stop_db 不再提前返回,现在返回正确的返回码。 |
| VER-45474 | Optimizer | DELETE and UPDATE now use optimized query plans accounting for down nodes. DELETE 和 UPDATE 现在使用考虑故障节点的优化查询计划。 |
| VER-68356 | Security | LDAPLink now properly trims whitespace from LDAPLinkSearchBase. LDAPLink 现在正确修剪 LDAPLinkSearchBase 中的空白字符。 |
| VER-70760 | Optimizer | Subqueries on tables with TopK projections no longer error. 对具有 TopK 投影的表的子查询不再报错。 |
| VER-71074 | Security | DROP privilege (plus CREATE on SCHEMA) is now sufficient to replace a view with CREATE OR REPLACE VIEW. DROP 权限(加上对 SCHEMA 的 CREATE 权限)现在足以使用 CREATE OR REPLACE VIEW 替换视图。 |
| VER-72232 | ComplexTypes | QUERY_EVENTS system table no longer recommends collecting stats on ARRAY & SET columns. QUERY_EVENTS 系统表不再推荐对 ARRAY 和 SET 列收集统计信息。 |
| VER-72394 | Execution Engine | Root operator now correctly detects whether a query is a DML plan, fixing row-count statistics. 根运算符现在正确检测查询是否为 DML 计划,修复了行计数统计。 |
| VER-72735 | Security | LDAPLink now checks for circular role/group assignments, logs the error, and skips the rest of the circular relationship. LDAPLink 现在检查循环角色/组分配,记录错误,并跳过剩余循环关系。 |
| VER-72850 | Optimizer | Incorrect usage of EXPORT_TABLES in EXPORT TO VERTICA now returns an error instead of crashing. 在 EXPORT TO VERTICA 中错误使用 EXPORT_TABLES 现在返回错误而不是崩溃。 |
| VER-72978 | Depot | Depot size changes via alter_location_size now propagate to all nodes in parallel.通过 alter_location_size 更改 depot 大小现在并行传播到所有节点。 |
| VER-73137 | Data Removal - Delete, Purge, Partitioning | ILM operations now obtain global catalog locks only after storage/migration is complete, reducing blocking. ILM 操作现在仅在存储/迁移完成后才获取全局目录锁,减少了阻塞。 |
| VER-73110 | Backup/DR | migrate_enterprise_to_eon now lists all tables with problematic projections in the error log.migrate_enterprise_to_eon 现在在错误日志中列出所有具有问题投影的表。 |
| VER-73382 | Execution Engine | Sort Order Violation errors no longer occur when join output sorted like the target projection is consumed by a single Analytic function. 当按目标投影排序的连接输出被单个 Analytic 函数消费时,不再出现 Sort Order Violation 错误。 |
| VER-73811 | Execution Engine | IN () expressions on LONG VARCHAR or LONG VARBINARY no longer cause slow execution or excessive memory usage. 对 LONG VARCHAR 或 LONG VARBINARY 的 IN () 表达式不再导致执行缓慢或内存使用过多。 |
| VER-73841 | Admin Tools | Adding nodes on large, busy databases now runs fewer queries and takes less time. 在大型繁忙数据库上添加节点现在运行更少的查询并花费更少的时间。 |
| VER-73881 | Scrutinize | Scrutinize now collects a memory report log. Scrutinize 现在收集内存报告日志。 |
| VER-73937 | Execution Engine | METADATA resource pool size is now calculated correctly. METADATA 资源池大小现在被正确计算。 |
| VER-74043 | Admin Tools | Checks of user storage locations during revive Eon database have been relaxed. 恢复 Eon 数据库期间对用户存储位置的检查已放宽。 |
| VER-74116 | Optimizer | Analytic functions with a LIMIT OVER clause no longer trigger an Optimizer error. 带有 LIMIT OVER 子句的分析函数不再触发优化器错误。 |
| VER-74130 | Depot, Tuple Mover | enforce_object_storage_policy() now returns a single error message stating it's not supported in Eon mode, instead of flooding logs.enforce_object_storage_policy() 现在返回一条错误消息,说明在 Eon 模式下不受支持,而不是淹没日志。 |
| VER-74182 | Admin Tools | Performance of db_remove_subcluster improved for subclusters with several nodes.db_remove_subcluster 对具有多个节点的子集群的性能已改进。 |
| VER-74300 | Backup/DR | migrate_enterprise_to_eon now excludes privileges on external procedures and checks for consistent dependencies.migrate_enterprise_to_eon 现在排除外部过程的权限并检查一致依赖关系。 |
| VER-74350 | Execution Engine | Performance for tables with >1600 columns has been addressed. 针对超过 1600 列的表性能问题已得到解决。 |
| VER-74355 | ComplexTypes | ALTER TABLE ... ALTER COLUMN now correctly errors on attempts to create unsupported multi-dimensional arrays. ALTER TABLE ... ALTER COLUMN 现在在尝试创建不支持的多维数组时正确报错。 |
| VER-74512 | Depot | DEPOT_PIN_POLICIES is now always empty in Enterprise mode. DEPOT_PIN_POLICIES 在 Enterprise 模式下现在始终为空。 |
| VER-74521 | Optimizer | INSERT...SELECT with LIMIT OVER no longer unnecessarily transfers data to the initiator. 带 LIMIT OVER 的 INSERT...SELECT 不再不必要地将数据传输到发起者节点。 |
| VER-74575 | Optimizer | GUA optimization now defers until after all query-related attributes are populated, fixing outer join + LIMIT OVER errors. GUA 优化现在推迟到所有查询相关属性填充之后执行,修复了外连接 + LIMIT OVER 错误。 |
| VER-74576 | Execution Engine | COUNT window aggregates over Fixed type columns including UUIDs no longer hit assertion failures. 对包括 UUID 在内的 Fixed 类型列的 COUNT 窗口聚合不再触发断言失败。 |
| VER-74588 | Hadoop | Queries to external tables with Hive partition columns no longer crash if a predicate on a partition column throws an error. 对具有 Hive 分区列的外部表的查询,如果分区列上的谓词抛出错误,不再崩溃。 |
| VER-74616 | Execution Engine | Inter-node encryption performance has been improved across various workloads. 跨各种工作负载的节点间加密性能已改进。 |
| VER-74707 | Tuple Mover | Mergeout caching is now auto-disabled when reflexive mergeout is enabled. 当自反 mergeout 启用时,Mergeout 缓存现在自动禁用。 |
| VER-74783 | DDL - Table | ALTER TABLE ... OWNER TO now returns NOTICE instead of ERROR when the table is already owned by the specified user. 当表已由指定用户拥有时,ALTER TABLE ... OWNER TO 现在返回 NOTICE 而不是 ERROR。 |
| VER-74795 | Optimizer | The optimizer now supports PROJS hints specifying live aggregate projections (LAP). 优化器现在支持指定实时聚合投影(LAP)的 PROJS 提示。 |
| VER-74918 | Security | LDAPLink now supports attributes with >1500 returned members. LDAPLink 现在支持返回超过 1500 个成员的属性。 |
| VER-75051 | Execution Engine | Simple CASE statements of the form CASE expression WHEN comparand THEN result ... ELSE result END no longer fail due to internal memory errors.形式为 CASE expression WHEN comparand THEN result ... ELSE result END 的简单 CASE 语句不再因内部内存错误而失败。 |
| VER-75105 | Execution Engine | Passing a volatile expression multiple times to an inline function now returns an error instead of incorrect output. 将易变表达式多次传递给内联函数现在返回错误而不是错误输出。 |
| VER-75196 | Execution Engine | ALTER TABLE REORGANIZE after setting group partitioning no longer crashes Vertica. 设置组分区后执行 ALTER TABLE REORGANIZE 不再导致 Vertica 崩溃。 |
| VER-75265 | Execution Engine | Subqueries in expressions implying correlated joins with aggregation no longer return unexpected multiple rows. 表达式中包含相关连接与聚合的子查询不再返回意外的多行结果。 |