docs: update roadmap (#34751)

Author: guanshengliang

docs/en/29-roadmap/index.md

@@ -6,7 +6,7 @@ The 2026 roadmap for TDengine OSS is described in the following table.
 
 |  Quarter   |  Feature  |
 | :----- | :----- |
-| 2026Q1 | <ol><li>Storage: batch tag modification, dynamic adjustment of last cache LRU</li><li>Query: subqueries, external windows, ANY/SOME/ALL/EXISTS operators, window and interpolation enhancements, Explain and Show Queries optimization</li><li>Virtual Tables: virtual tables support referencing virtual tables, virtual table query performance optimization, subscription to virtual table metadata changes</li><li>Stream: triggering by natural week/month/quarter/year, event trigger condition optimization, grouped computation performance optimization, virtual supertable triggering supports child table additions/deletions/modifications</li></ol> |
-| 2026Q2 | <ol><li>Storage: multi-replica leader switching and node recovery optimization, last cache optimization (such as forced refresh, multi-column write optimization, specified column caching)</li><li>Query: join queries, relaxed window query restrictions, time windows support week/month/quarter/year, Interp performance improvement</li><li>Functions: window functions and OVER clause</li><li>Virtual Tables: support for virtual table inheritance, virtual supertable column name modification</li><li>Stream: performance optimization for multiple customer scenarios, performance optimization for multi-measurement point scenarios, historical computation performance optimization, virtual table triggering performance optimization</li><li>Others: TDgpt predictive maintenance, minute-level timezone support</li></ol> |
-| 2026Q3 | <ol><li>Query: complex query performance optimization, scalar correlated subqueries, cumulative windows, further enhancement of join queries, SQL testing tools</li><li>Functions: 15 customer-requested functions</li><li>Virtual Tables: subscription to virtual table data through continuous queries</li><li>Stream: maintainability improvement</li><li>Others: TDgpt supports PCA, PLS, cluster analysis</li></ol> |
-| 2026Q4 | <ol><li>TEXT data type, variable-length string data type, database name modification, column name modification</li><li>Query: query parallelization, addition of observability metrics</li><li>Functions: support for MySQL operators and functions, UDF function framework refactoring</li><li>Others: engine CPU and memory management</li></ol> |
+| 2026Q1 | <ol><li>Storage: batch tag modification, dynamic adjustment of last cache LRU</li><li>Query: subqueries, external windows, ANY/SOME/ALL/EXISTS operators, window and interpolation enhancements, Explain and ShowQueries optimization</li><li>Virtual Tables: virtual table query performance optimization, subscription to virtual table metadata changes, validation of references between virtual tables and source tables</li><li>Stream: triggering by natural week/month/quarter/year, true_for conditions for event and state window triggers support duration and count, grouped computation performance optimization, virtual supertable triggering supports child table additions/deletions/modifications</li></ol> |
+| 2026Q2 | <ol><li>Storage: multi-replica leader switching and node recovery optimization, last cache optimization (such as forced refresh, multi-column write optimization, specified column caching)</li><li>Query: federated query demo version, relaxed window query restrictions, time windows support week/month/quarter/year, Interp performance improvement</li><li>Functions: window functions and OVER clause</li><li>Virtual Tables: virtual table inheritance, virtual tables referencing virtual tables, virtual supertable column name modification</li><li>Stream: performance optimization for multiple customer scenarios, performance optimization for multi-measurement point scenarios, historical computation performance optimization, virtual table triggering performance optimization</li><li>TDgpt: model lifecycle management, predictive maintenance, minute-level timezone support</li></ol> |
+| 2026Q3 | <ol><li>Storage: improved data splitting and migration performance, data reorganization without affecting writes, disk rate limiting</li><li>Query: official release of federated queries, complex query performance optimization, scalar correlated subqueries, cumulative windows, SQL testing tools</li><li>Functions: 15 customer-requested functions</li><li>Virtual Tables: subscription to virtual table data through continuous queries</li><li>Stream: maintainability improvement</li><li>Others: TDgpt supports PCA, PLS, cluster analysis</li></ol> |
+| 2026Q4 | <ol><li>Storage: full and incremental backup, TEXT data type, variable-length string data type, database name modification, column name modification</li><li>Query: further enhancement of join queries, query parallelization, addition of observability metrics</li><li>Functions: support for MySQL operators and functions, UDF function framework refactoring</li><li>Others: engine CPU and memory management</li></ol> |

docs/zh/29-roadmap/index.md

@@ -6,7 +6,7 @@ TDengine TSDB OSS 之 2026 年年度路线图如下表所示。
 
 |  季度   |  功能  |
 | :----- | :----- |
-| 2026Q1 | <ol><li>存储：批量标签修改、动态调整数据缓存的 LRU</li><li>查询：子查询、外部窗口、ANY/SOME/ALL/EXISTS 运算符、窗口及插值增强、Explain 和 Show Queries 优化</li><li>虚拟表：虚拟表支持引用虚拟表、虚拟表查询性能优化、订阅虚拟表的元数据变更</li><li>流计算：按自然周/月/季/年触发、事件触发条件优化、分组计算性能优化、虚拟超级表触发支持子表增删改</li></ol> |
-| 2026Q2 | <ol><li>存储：多副本切主和节点恢复优化、数据缓存优化（如强制刷新、多列写入优化、指定列缓存）</li><li>查询：关联查询，放宽窗口查询限制、时间窗口支持周/月/季/年、Interp 性能提升</li><li>函数：窗口函数及 OVER 字句</li><li>虚拟表：支持虚拟表继承、虚拟超级表列名修改</li><li>流计算：多个客户场景性能优化、多测点场景性能优化、历史计算性能优化、虚拟表触发性能优化</li><li>其他：TDgpt 预测性维护、分钟级时区</li></ol> |
-| 2026Q3 | <ol><li>查询：复杂查询性能优化、标量相关子查询、累计窗口、关联查询进一步增强、SQL 测试工具</li><li>函数：15 个客户需要的函数</li><li>虚拟表：通过连续查询订阅虚拟表数据</li><li>流计算：可维护性提升</li><li>其他：TDgpt 支持 PCA、PLS、聚类分析</li></ol> |
-| 2026Q4 | <ol><li>存储：TEXT 数据类型、不定长字符串数据类型、库名修改、列名修改</li><li>查询：查询并行化、增加可观测指标</li><li>函数：支持 MySQL 运算符及函数、UDF 函数框架重构</li><li>其他：引擎的 CPU 和内存管控</li></ol> |
+| 2026Q1 | <ol><li>存储：批量标签修改、动态调整数据缓存的 LRU</li><li>查询：子查询、外部窗口、ANY/SOME/ALL/EXISTS 运算符、窗口及插值增强、Explain 和 Show Queries 优化</li><li>虚拟表：虚拟超级表查询性能优化、订阅虚拟表的元数据变更、虚拟表和源表的引用校验</li><li>流计算：按自然周/月/季/年触发、事件和状态窗口触发的 true_for 条件支持持续时间与条数、分组计算性能优化、虚拟超级表触发支持子表增删改</li></ol> |
+| 2026Q2 | <ol><li>存储：多副本切主和节点恢复优化、数据缓存优化（如强制刷新、多列写入优化、指定列缓存）</li><li>查询：联合查询演示版本，放宽窗口查询限制、时间窗口支持周/月/季/年、Interp 性能提升</li><li>函数：窗口函数及 OVER 字句</li><li>虚拟表：虚拟表继承、虚拟表引用虚拟表、虚拟超级表列名修改</li><li>流计算：多个客户场景性能优化、多测点场景性能优化、历史计算性能优化、虚拟表触发性能优化</li><li>TDgpt：模型生命周期管理、预测性维护、分钟级时区</li></ol> |
+| 2026Q3 | <ol><li>存储：提升数据拆分和移动的性能、数据重整不影响写入、磁盘限速</li><li>查询：联合查询正式发布、复杂查询性能优化、标量相关子查询、累计窗口、SQL 测试工具</li><li>函数：15 个客户需要的函数</li><li>虚拟表：通过连续查询订阅虚拟表数据</li><li>流计算：可维护性提升</li><li>其他：TDgpt 支持 PCA、PLS、聚类分析</li></ol> |
+| 2026Q4 | <ol><li>存储：全量与增量备份、TEXT 数据类型、不定长字符串数据类型、库名修改、列名修改</li><li>查询：关联查询进一步增强、查询并行化、增加可观测指标</li><li>函数：支持 MySQL 运算符及函数、UDF 函数框架重构</li><li>其他：引擎的 CPU 和内存管控</li></ol> |