26.2.x Resolved issues¶
26.2.0-0¶
Updated 4/29/2026
| Issue Key | Application Module | Description |
|---|---|---|
| 4234407 | Kubernetes | Enhanced Vertica Database Initialization for Custom Setups: Previously, coordinating initial setup tasks for Vertica databases on Kubernetes pods, such as creating users, schemas, or installing UDX libraries, was challenging. Users often resorted to complex workarounds to ensure these configurations ran only after the database was fully initialized. Solution: We have introduced a new capability that allows you to define a custom job specification within VerticaDB. This job will automatically execute a user-defined container once the Vertica database has been successfully initialized. Benefits: This enhancement simplifies and automates the process of performing critical post-initialization setups. Customers can now seamlessly integrate custom configurations, run Ansible scripts, or deploy C++ UDX libraries, ensuring their Vertica environment is ready for use without manual intervention or complex coordination logic. 在 Kubernetes 上自定义初始化 Vertica 数据库的能力增强。新增在 VerticaDB 中定义自定义 Job 规格的功能，数据库初始化完成后自动执行用户定义的容器，用于创建用户、Schema、安装 UDX 库等任务，无需复杂编排。 |
| 4278598 | Logging | Improved Log Rotation for GCLX Profile Files: Previously, GCLX profile (.prof) files located in catalog_dir/DebugReports/ directories could accumulate indefinitely, leading to excessive disk space consumption. This update integrates these files into the existing in-server log rotation mechanism, ensuring they are properly cleaned up over time. This enhancement helps maintain system health and prevents potential customer annoyance due to unmanaged file growth.改进了 GCLX Profile ( .prof) 文件的日志轮转机制。之前这些文件会无限累积导致磁盘空间耗尽，现已集成到服务端日志轮转中，自动清理过期文件。 |
| 4340995 | Client Drivers - Misc | ODBC-Loader: PCRE Library Compatibility on Amazon Linux and RHEL 8.x: The ODBC-Loader previously encountered compatibility issues on Amazon Linux and Red Hat Enterprise Linux 8.x environments. This was due to a dependency on a specific build of the PCRE (Perl Compatible Regular Expressions) library, which led to runtime errors (missing symbols from libpcrecpp) when standard system-provided PCRE packages were utilized. Resolution/Workaround: To ensure the correct functionality of the ODBC-Loader in these specific environments, a manual procedure is required. Users must perform the following steps: 1. Remove any existing pcre-devel and pcre-cpp packages. 2. Manually download and compile PCRE version 8.45 from its source, ensuring the use of specific compiler flags (CXXFLAGS='-std=c++11 -D_GLIBCXX_USE_CXX11_ABI=0'). 3. Update the system's dynamic linker configuration to recognize the newly installed PCRE libraries. 4. Recompile the ODBC-Loader itself, applying the same specified CXXFLAGS used during the PCRE compilation. This detailed procedure ensures that the ODBC-Loader correctly links against a compatible PCRE build, thereby resolving the previously observed runtime issues.修复了 ODBC-Loader 在 Amazon Linux 和 RHEL 8.x 上的 PCRE 库兼容性问题。需要手动下载编译 PCRE 8.45 并使用特定编译器标志重新编译 ODBC-Loader 以解决运行时符号缺失问题。 |
| 4733747 | Execution Engine | Increased RemoteInitiatorBufSize Limit: Previously, users experienced query failures and performance issues due to an insufficient maximum limit for the RemoteInitiatorBufSize parameter, leading to "Attempted to return result set too large" errors. We have addressed this limitation by increasing the configurable maximum value for RemoteInitiatorBufSize. This enhancement allows users to set the buffer size to a sufficiently high value, preventing query failures for large result sets without negatively impacting overall query performance.提高了 RemoteInitiatorBufSize 参数的最大可配置值，解决了因缓冲区上限不足导致大结果集查询失败报 "Attempted to return result set too large" 错误的问题。 |
| 4846351 | Optimizer | In some cases analyze statistics could encounter internal type mismatches when processing projections causing an Optimizer assert failure. The cause of these mismatches has been resolved. 修复了统计信息分析时处理投影可能遇到内部类型不匹配导致 Optimizer 断言失败的问题。 |
| 4861270 | Optimizer | An "Internal Optimizer Error (11)", specifically related to isInputTypeCompatibleWithExpected(intype, vartype), has been resolved. This error previously occurred during the optimization of certain complex SQL queries involving Common Table Expressions (CTEs) and UNION operations, preventing successful execution.修复了涉及 CTE 和 UNION 的复杂 SQL 查询优化时出现的 "Internal Optimizer Error (11)" 内部优化器错误。 |
| 4892839 | EON | Improved run_index_tool for Communal Location Checks: The run_index_tool's CRC checker has been enhanced to more thoroughly inspect communal locations. Previously, the tool would skip unreadable files, potentially masking data corruption. With this improvement, the tool now accurately identifies and reports these unreadable files, along with their associated tables and projections. This allows customers to proactively detect corrupted data, facilitating recovery efforts and strengthening Vertica's data integrity validation mechanisms, including the --validate/-v option.增强了 run_index_tool 的 CRC 检查器对公共存储位置的检查能力。之前工具会跳过不可读文件可能掩盖数据损坏，现在能准确识别和报告这些文件及其关联表和投影。 |
| 4926506 | Optimizer | This release includes a fix for an issue where queries in Vertica 23.4 environments would intermittently fail with an "Internal Optimizer Error (11): PlanExprFixup: No mapping found" message. This error was observed particularly after migrating table data from Vertica 8.1.1 and executing queries that functioned correctly in the previous version. This error prevented queries from completing successfully, leading to disruptions in data operations and analysis.This issue has been resolved by addressing the root cause of the optimizer error. 修复了 Vertica 23.4 环境中查询间歇性失败报 "Internal Optimizer Error (11): PlanExprFixup: No mapping found" 的问题，此错误在从 8.1.1 迁移表数据后尤其容易出现。 |
| 4937780 | Execution Engine | An enhancement was added in version 24.1 which preserved the outer's sortedness for hash/merge right outer joins, if it was compatible. With this ticket, we saw that it is possible in a merge join for that enhancement to result in a sort violation (which can create inconsistent results with aggregate functions, as in this case). In order to maintain the sort, we added back the explicit Sort path when we're dealing with a merge join. 修复了 24.1 版本中为保持外测排序性而添加的增强功能在 merge join 中可能导致排序违规的问题。现在在处理 merge join 时重新添加了显式 Sort 路径。 |
| 4967176 | UDX | Java UDx JVM Sharing for Enhanced Efficiency: Previously, each session executing Java User-Defined Extensions (UDx) would instantiate its own Java Virtual Machine (JVM). This update introduces the capability for multiple user sessions to share a single Java UDx JVM. This enhancement significantly reduces memory overhead, lowers resource costs, and improves system scalability for Java UDx workloads. Java UDx JVM 共享能力增强。之前每个执行 Java UDx 的会话都会启动独立的 JVM，现在多个用户会话可以共享同一个 JVM，显著降低内存开销和资源成本。 |
| 5000996 | Optimizer | Previously, nodes could be brought down by executing a MERGE query when a target value of the INSERT statement was a subquery. Now, a proper error message is displayed when this is attempted which is in line with the restrictions for using a subquery as a value for an INSERT statement outside of a MERGE query. 修复了 MERGE 查询中 INSERT 语句的目标值为子查询时可能导致节点宕机的问题。现在会显示正确的错误信息，与 MERGE 外部使用子查询的限制保持一致。 |
| 5061268 | Execution Engine | Previously, when a user using the COPY command with the delimited parser was trying to insert data for an array[varchar] column, they could trigger an internal error if their input began with back-to-back enclosing characters. Now, the server appropriately can parse multiple enclosing characters to begin an array[varchar] column, where the first enclosing character is treated as an enclosing character and the second enclosing character represents the first character in the varchar data. 修复了使用 COPY 命令和 delimited 解析器向 array[varchar] 列插入数据时，如果输入以连续封闭字符开头可能触发内部错误的问题。 |
| 5086469 | Vcluster | Improved VCluster Configuration for AWS Access: This update addresses customer feedback regarding the management of AWS access information within VCluster. Previously, users were required to provide AWS access details in a separate file, which caused inconvenience. We have now streamlined this process by migrating these critical configuration parameters into config_param.json. This change simplifies VCluster setup and enhances the overall user experience by centralizing access information.改进了 VCluster 中 AWS 访问信息的配置方式。之前需要单独文件提供 AWS 凭据，现在将这些关键配置参数迁移到 config_param.json 中集中管理，简化了设置流程。 |
| 5106995 | DDL - Table | Previously, a scheduled stored procedure would execute with a different default value for standardConformingStrings. This would produce differing results than a stored procedure executing using the CALL functionality if the stored procedure query contained usage of backslash impacted by the setting of standardConformingStrings. Now, it is aligned with the same default value as user sessions.修复了定时存储过程与 CALL 方式执行存储过程在 standardConformingStrings 默认值上不一致的问题，现在两者使用相同的默认值，确保执行结果一致。 |
| 5107285 | Optimizer | Query Performance Degradation with Views Containing Null Columns: Fixed an issue where queries utilizing views that included additional null columns could lead to unexpected changes in the query execution plan. This often resulted in significantly higher memory consumption and 'join inner did not fit into the memory' errors, particularly impacting some standard queries. This fix ensures more stable and optimized query execution, preventing out-of-memory issues and improving the reliability and performance of queries involving such views. 修复了使用包含 null 列的视图可能导致查询计划异常变化的问题。之前这会导致内存消耗激增和 "join inner did not fit into the memory" 错误。 |
| 5112215 | DDL - Table | This fix addresses the root cause allowing non-UTF-8 data and oversized byte strings to be stored in VARCHAR columns. The system now correctly enforces VARCHAR length constraints and ensures proper UTF-8 character handling during data insertion and processing, especially within recursive CTEs and temporary relation reuse mechanisms. This prevents the storage of invalid data and eliminates the associated query failures. Workarounds (Prior to Fix): Before the implementation of this fix, customers could mitigate the issue by: 1. Disabling the EnableWITHTempRelReuseLimit session parameter. 2. Enabling the WithClauseMaterialization session parameter. However, this option was noted to cause significant performance degradation, particularly with higher WithClauseRecursionLimit values.修复了 VARCHAR 列中可能存储非 UTF-8 数据和超长字节串的问题。现在系统正确执行 VARCHAR 长度约束，确保数据插入和处理过程中的 UTF-8 字符正确处理。 |
| 5118476 | Execution Engine | Excessive vertica.log Growth Leading to Node Failures: Previously, an issue was identified where vertica.log files would grow excessively large, sometimes by over 20GB in 30 minutes, due to frequent logging of [EE] <INFO> arg1, arg2, res: messages. This uncontrolled log growth consumed all available disk space in the catalog directory, leading to node failures and impacting subcluster availability. Resolution: This fix addresses the root cause of the excessive logging. The system has been updated to prevent the frequent generation of [EE] <INFO> arg1, arg2, res: messages at an INFO level, thereby controlling the size of vertica.log files. Impact: This resolution prevents future node failures caused by disk space exhaustion due to log growth, ensuring greater system stability and availability for Eon mode deployments running Vertica v24.3.0-1 and later.修复了 vertica.log 文件因频繁记录 [EE] <INFO> arg1, arg2, res: 消息而过度增长（30 分钟内增长超过 20GB）导致节点故障的问题。 |
| 5119497 | Catalog Engine | Concurrent sequence generation in OCC lock mode can now tolerate LockTimeout setting to zero. Instead of immediate failing, they now wait for the maximum amount of time given by LockTimeout and a new knob called MinOCCLockTimeout.并发序列生成在 OCC 锁模式下现在可以容忍 LockTimeout 设置为零。不再立即失败，而是等待 LockTimeout 和新增的 MinOCCLockTimeout 参数共同决定的最大等待时间。 |
| 5130566 | Data load / COPY | Iceberg Manifest File Handling Improvement: Resolved an issue where Iceberg manifest Avro files, particularly those generated by DataBricks Unity Catalog, could cause application crashes when the value_counts property was present with a null value. The system now correctly processes these null values, enhancing stability and compatibility with various Iceberg data sources.修复了 Iceberg manifest Avro 文件中 value_counts 属性为 null 值时可能导致应用崩溃的问题，增强了对 DataBricks Unity Catalog 等 Iceberg 数据源的兼容性。 |
| 5139801, 5139802 | Security | This release includes important security updates to address identified critical and high-severity vulnerabilities within the vertica-23.4.0-21.x86_64.SUSE.rpm component. These updates ensure compliance with release criteria requiring zero critical and high security findings. Users are advised to upgrade to this version to benefit from enhanced security posture.此版本包含了针对 vertica-23.4.0-21.x86_64.SUSE.rpm 组件中识别出的关键和高严重性漏洞的重要安全更新。 |
| 5142431 | EON | Vertica Eon Mode - Corrected TEMP Data Location for Newly Added Nodes: Previously, in Vertica Eon mode clusters, when new nodes were added following a database revive operation, the DATA,TEMP storage locations on these newly provisioned nodes were incorrectly configured to point to the catalog path (e.g., /vertica_catalog) instead of the intended data path (e.g., /vertica_tmp or the specified data directory). This behavior deviated from the standard configuration observed on original nodes or nodes added to a freshly created database. Resolution: This issue has been resolved. With this fix, all newly added nodes to an Eon mode cluster, even after a database revive, will now correctly configure their DATA,TEMP storage locations to the designated data paths, ensuring consistent and proper storage management across the cluster.修复了 Eon 模式下数据库恢复操作后新增节点的 DATA,TEMP 存储位置错误指向 catalog 路径而非指定数据路径的问题，确保存储管理一致性。 |
| 5144077 | Execution Engine | This fix addresses "Insufficient resources to execute plan on pool general" errors, specifically memory exhaustion, causing transactions to fail. This issue affected queries that previously ran without problems on Vertica 10.1v, even after increasing node RAM from 500GB to 600GB. This could lead to operational disruptions and resource allocation problems within the Vertica cluster. Resolution: This fix resolves the memory resource allocation inefficiencies and errors, enabling successful query execution within configured resource limits. 修复了 "Insufficient resources to execute plan on pool general" 内存耗尽错误，解决了 Vertica 10.1 版本可正常运行的查询在更高版本中即使增加内存也无法执行的问题。 |
| 5158893 | EON | Catalog Truncation for Compressed Transaction Logs Issue: Previously, when catalog truncation was performed on compressed transaction log files, an incorrect counter was used to determine the log size. This led to the transaction log file being truncated more than necessary, resulting in corruption and subsequent startup failures. Resolution: The system now correctly identifies and uses the appropriate counter to determine the size of compressed transaction log files during truncation. This ensures that the last transaction file is truncated accurately, preventing corruption and ensuring successful restarts. 修复了压缩事务日志文件在执行 catalog truncation 时使用错误计数器确定日志大小，导致过度截断进而损坏日志并引发启动失败的问题。 |
| 5162622 | Security | Fix for FAILED_LOGIN_ATTEMPTS Profile Parameter: Previously, the FAILED_LOGIN_ATTEMPTS profile parameter in Vertica did not function as expected. When a user profile was configured with a specific limit for failed login attempts (e.g., 5 attempts), the user account would unexpectedly lock after a fewer number of failed attempts (e.g., 3 attempts). This discrepancy led to premature account lockouts, causing inconvenience for end-users and impacting business operations. This behavior was specifically observed and reproduced in Vertica Analytic Database v25.1.0-0. Resolution: This issue has been resolved. The FAILED_LOGIN_ATTEMPTS profile parameter now correctly enforces the configured number of failed login attempts before locking a user account. Users will no longer experience premature account lockouts due to this parameter, ensuring that security policies are applied as intended.修复了 FAILED_LOGIN_ATTEMPTS 配置参数功能异常的问题。之前配置了失败尝试次数限制后，用户账户会在更少的尝试次数后意外锁定。 |
| 5162699 | Installation Program | --ssh-password Option Failure in install_vertica and update_vertica: Previously, the --ssh-password option in the install_vertica and update_vertica scripts was not being correctly applied. This led to the scripts prompting for an SSH password and becoming unresponsive during installation or upgrade processes, especially in environments where passwordless SSH was not configured. This behavior resulted in incomplete upgrades and operational disruptions. Resolution: The --ssh-password option is now properly recognized. The install_vertica and update_vertica scripts now correctly utilize the provided SSH password, ensuring smooth and uninterrupted execution without unexpected password prompts.修复了 install_vertica 和 update_vertica 脚本中 --ssh-password 选项未被正确应用的问题，解决了脚本在无密码 SSH 环境中挂起的问题。 |
| 5163573 | S3 | S3 Streaming Connection Management Improvement: Previously, the S3 layer could improperly count streaming connections, potentially leading to connection pool exhaustion during surges of S3 requests. This issue could block S3 reads, including critical operations like TxnLogSync, and impact query performance. This release includes an improvement to ensure streaming connections are counted atomically and managed correctly, preventing connection pool exhaustion and ensuring reliable S3 read operations, even under high load.改进了 S3 流式连接计数管理。之前可能错误计数导致连接池在高 S3 请求流量下耗尽，阻塞 S3 读取操作。现在确保连接原子计数和正确管理。 |
| 5174220 | Kafka Integration | Core Vulnerability (CVE-2025-68161): This update addresses a critical security vulnerability, CVE-2025-68161, found in org.apache.logging.log4j_log4j-core version 2.17.1. The vulnerability was identified during Prisma scans on container_minimal_arm environments. The Log4j core library has been upgraded to version 2.25.3 to mitigate this issue.修复了 org.apache.logging.log4j_log4j-core 2.17.1 版本中的严重安全漏洞 CVE-2025-68161，已将 Log4j 核心库升级到 2.25.3。 |
| 5182302 | Admin Tools | Before this release, AdminTools restart_node may find a secondary UP node to run catalog editor for getting catalog information. After this fix, AdminTools restart_node consistently finds a primary UP node to run catalog editor for getting catalog information. This ensures you're getting the correct and up-to-date catalog information.修复了 AdminTools restart_node 可能选择次要 UP 节点运行 catalog editor 的问题。现在一致地选择主 UP 节点，确保获取正确和最新的 catalog 信息。 |
| 5195379 | Security | Vertica TLS Key File Path Recognition: An issue has been resolved where Vertica version 25.1.0-3 incorrectly recognized the TLS key file location as "/.vsql/" instead of the expected "/home/vertica/.vsql/". This prevented successful TLS connections, particularly for customers upgrading from version 24.4.0-1. The fix ensures that Vertica now correctly identifies the TLS key file path, allowing for proper TLS connectivity. 修复了 Vertica 25.1.0-3 错误识别 TLS 密钥文件路径为 /.vsql/ 而非期望的 /home/vertica/.vsql/ 的问题，确保 TLS 连接正常。 |
| 5195776 | Client Drivers - ODBC | Vertica ODBC Driver Performance Restoration for Data Unloads: This fix addresses a significant performance degradation observed in newer Vertica ODBC drivers (versions 23.04.0000 and 25.01.0005). Previously, customers utilizing SAS software for data transformations and reporting experienced a 10-20% reduction in unload performance when migrating from older driver versions (e.g., 11.00.0200). This update ensures that data unload operations now perform at expected speeds, restoring efficiency for users. 修复了新版本 ODBC 驱动（23.04.0000 和 25.01.0005）在数据卸载场景下的性能退化问题（比旧版本 11.00.0200 慢 10-20%），恢复预期的卸载性能。 |
| 5196319 | EON | A new internal knob called RemoveBadRestorePointsOnStartup is added to allow progress of startup in the presence of bad restore points, if enabled.新增内部参数 RemoveBadRestorePointsOnStartup，启用后允许在存在损坏恢复点时继续启动，避免启动阻塞。 |
| 5206178 | Node Management Agent | Security Fix: Golang Vulnerabilities in Vertica Binaries: This release addresses multiple security vulnerabilities found in Golang binaries shipped with Vertica 26.1.0-0. These vulnerabilities, including several critical and high-severity CVEs, affected core Vertica components and could prevent deployment in Kubernetes environments with strict vulnerability-based admission controllers. Key Vulnerabilities Addressed: - CVE-2024-45337 (CRITICAL): Authorization bypass in golang.org/x/crypto/ssh. - CVE-2025-68121 (CRITICAL/UNKNOWN): Session resumption issue in crypto/tls. - CVE-2025-22869 (HIGH): Denial of Service in golang.org/x/crypto/ssh. - CVE-2025-61726 (HIGH): Memory exhaustion in net/url query parameter parsing. Resolution: The fix involves updating vulnerable Golang dependencies, specifically golang.org/x/crypto to v0.45.0, golang.org/x/ne t to v0.47.0, and github.com/go-viper/mapstructure/v2 to v2.4.0. Additionally, the Go compiler has been upgraded to version 1.24.13 to resolve stdlib vulnerabilities. This update significantly enhances the security posture of Vertica container images. Affected Components: - opt/vertica/bin/vcluster - opt/vertica/bin/vcluster_server - opt/vertica/bin/vertica-client-proxy - opt/vertica/bin/vertica_mcp_server - opt/vertica/bin/node_management_agent修复了 Vertica 26.1.0-0 中 Golang 二进制文件的多项安全漏洞，包括 crypto/tls 中的严重 CVE-2025-68121 等，升级了 Go 编译器到 1.24.13。 |
| 5206299 | Execution Engine | Database Crash with ~ ANY (Subquery): Fixed a critical issue that caused the Vertica database to crash unexpectedly with a SIGSEGV error when processing queries containing the ~ ANY (Subquery) construct. This fix improves database stability and prevents service unavailability.修复了处理包含 ~ ANY (Subquery) 结构的查询时导致 Vertica 数据库因 SIGSEGV 意外崩溃的关键问题。 |
| 5208747 | Kubernetes, Vcluster | Before the fix, the Kubernetes operator may error out upon seeing remote user storage location when reviving an Eon database migrated from 10.x Enterprise database. With the fix, the Kubernetes operator correctly handles the remote user storage locations created in old version enterprise database for reviving. 修复了 Kubernetes Operator 在恢复从 10.x Enterprise 数据库迁移的 Eon 数据库时遇到远程用户存储位置报错的问题。 |
| 5222928 | Security | Enhanced Security: ECC Support for SSL Certificates: Vertica now supports Elliptic Curve Cryptography (ECC) for SSL certificates, offering a more modern and robust alternative to the traditional RSA algorithm. This enhancement significantly strengthens Vertica's SSL/TLS security posture, aligning with contemporary security best practices and providing a more secure communication environment. Vertica 现在支持 SSL 证书的椭圆曲线加密（ECC），提供了比传统 RSA 算法更现代、更健壮的替代方案，增强了 SSL/TLS 安全防护能力。 |
| 5224783 | Backup/DR | Performance Improvement for Snapshot Loading: This release addresses a performance bottleneck observed during the load_snapshot() operation, which could sometimes take up to 4 hours on certain nodes. Key Improvements: - Optimized Storage Container Checks: Reduced redundant checks for storage locations within Storage Containers (SCs). - Streamlined CRC Verification: Introduced optimizations to skip unnecessary CRC checks for existing storage bundles. These enhancements significantly improve the efficiency and reduce the duration of the load_snapshot() process, leading to faster system recovery and improved overall performance.优化了 load_snapshot() 操作，之前在某些节点上可能耗时长达 4 小时。通过减少存储容器冗余检查和优化 CRC 验证，显著提升快照加载效率。 |
| 5225942 | Complex Types | View Creation with UNNEST and CASE Expressions: Previously, users encountered an ERROR 3457: Function unnest(-) does not exist when attempting to create a view that utilized the unnest function within a CASE expression. This specific issue manifested when the CASE expression involved NULL arrays of complex types, such as ARRAY[ROW(...)] or ARRAY[ARRAY[...]]. Resolution: This release includes a fix that resolves the type resolution issue for unnest functions within CASE expressions during view creation. Users can now successfully define views containing unnest with CASE expressions, even when handling NULL arrays of complex types, without encountering the aforementioned error.修复了在 CASE 表达式中使用 unnest 函数创建视图时报 ERROR 3457: Function unnest(-) does not exist 的问题，涉及复杂类型的 NULL 数组场景。 |
| 5254733 | Admin Tools | Critical and High Security Vulnerabilities Addressed: This release includes important security updates to address several critical and high-severity vulnerabilities (CVEs) identified in version 23.4.0-23. These fixes enhance the overall security and stability of the system. 此版本包含了针对 23.4.0-23 版本中识别出的多个关键和高严重性安全漏洞（CVE）的重要安全更新。 |
| 5270913 | Data load / COPY | FCSVPARSER used to incorrectly parse CRLF record terminators leading in some cases to reject the data value of the last field in the record. Now, CRLF record terminators are properly parsed. 修复了 FCSVPARSER 在某些情况下错误解析 CRLF 记录终止符，导致记录中最后一个字段数据被拒绝的问题。 |
| 5280338 | Optimizer | Overly restrictive error handling was added for complex/nested complex types containing user-defined functions. This was causing some valid SQL to report an 'unsupported' error when it should not have. This issue has been fixed. 修复了对包含用户自定义函数的复杂/嵌套复杂类型过度限制的错误处理，之前导致一些有效的 SQL 被错误报告为 "unsupported"。 |
| 5293124 | Kubernetes | Security Update: verticadb-operator: This release includes an important security fix for verticadb-operator:26.1.2-0. Resolved Issue: - CVE-2026-33186: Addressed a critical security vulnerability in the gRPC component, which required an upgrade to a patched version. This update enhances the security posture of the operator.修复了 verticadb-operator:26.1.2-0 中的安全漏洞 CVE-2026-33186（gRPC 组件），升级到已修补的版本。 |
| 5297569 | Kubernetes | Scrutinize System Tables Collection on Kubernetes: This release addresses an issue where running Scrutinize using a Custom Resource Definition (CRD) on Kubernetes, specifically in prod and dev EON environments, failed to collect system tables. Previously, the generated Scrutinize tarball did not include these critical system tables, impacting diagnostic capabilities. With this fix, Scrutinize tarballs generated via CRD on Kubernetes will now correctly include all expected system tables, ensuring comprehensive data collection for troubleshooting and analysis. 修复了在 Kubernetes 上使用 CRD 运行 Scrutinize 时未能收集系统表的问题，现在生成的 Scrutinize 归档文件将正确包含所有系统表。 |
| 5326993 | Client Drivers - ODBC, Client Drivers - VSQL | Vertica 26.x Client Driver SSL Connectivity Fix Issue Addressed: This release resolves a critical defect affecting Vertica 26.1.0-3 client drivers (ODBC and vsql). Previously, these drivers failed to establish connections to Vertica servers with SSL enabled. Users experienced issues such as the ODBC test connection screen disappearing without a message and vsql exiting without an error, blocking customer operations. This behavior was specific to Vertica 26.x drivers, while Vertica 25.x drivers functioned correctly under the same conditions. Resolution: This fix ensures that Vertica 26.x client drivers now correctly connect to SSL-enabled Vertica servers, restoring stable and secure connectivity and aligning with the expected behavior observed in previous driver versions. 修复了 Vertica 26.x 客户端驱动（ODBC 和 vsql）无法连接到启用 SSL 的 Vertica 服务器的关键缺陷，确保 26.x 驱动与 SSL 服务器的正常连接。 |
| 5329625 | Backup/DR | Tombstone objects can have empty locations if queued after an upgrade from version 23.4 and older. For these, we now use the default storage location to avoid possible SIGSEGV. 修复了从 23.4 及更早版本升级后 tombstone 对象可能具有空位置的问题。现在使用默认存储位置以避免可能的 SIGSEGV。 |
| 5330533 | Client Drivers - JDBC | Vertica JDBC Hang on Upgraded Clusters: Resolved a critical issue where Vertica JDBC clients would intermittently hang or queries would fail with cancellation errors (SQL State 57014) on upgraded Vertica clusters (from 12.0.4-11 to 25.1 or 25.4). This problem primarily affected non-dbadmin users and, in some instances, all users on fresh Vertica 25.x installations, causing indefinite client hangs despite successful query execution on the server side. 修复了升级后的 Vertica 集群上 JDBC 客户端间歇性挂起或查询因取消错误（SQL State 57014）失败的问题，主要影响非 dbadmin 用户。 |
| 5332431 | UI - Management Console | Vertica MC Server: Login Session Stability Improvement: An issue has been fixed where users upgrading to version 26.1.0-3 of the Vertica Management Console experienced "key clock application session has expired" errors, preventing successful login. This fix ensures stable user sessions post-upgrade. 修复了升级到 Management Console 26.1.0-3 后用户遇到 "key clock application session has expired" 错误无法登录的问题，确保升级后用户会话稳定。 |